US20200058088A1 - Method and system for determining content treatment - Google Patents
Method and system for determining content treatment Download PDFInfo
- Publication number
- US20200058088A1 US20200058088A1 US16/360,896 US201916360896A US2020058088A1 US 20200058088 A1 US20200058088 A1 US 20200058088A1 US 201916360896 A US201916360896 A US 201916360896A US 2020058088 A1 US2020058088 A1 US 2020058088A1
- Authority
- US
- United States
- Prior art keywords
- content
- reference work
- uploaded
- uploaded content
- determining
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Abandoned
Links
- 238000000034 method Methods 0.000 title claims description 72
- 238000009826 distribution Methods 0.000 claims abstract description 26
- 230000009471 action Effects 0.000 claims description 22
- 230000000007 visual effect Effects 0.000 claims description 5
- 238000013459 approach Methods 0.000 abstract description 4
- 238000005516 engineering process Methods 0.000 description 28
- 230000008569 process Effects 0.000 description 11
- 239000011435 rock Substances 0.000 description 10
- 238000004422 calculation algorithm Methods 0.000 description 8
- 238000012545 processing Methods 0.000 description 8
- 230000001815 facial effect Effects 0.000 description 5
- 230000006855 networking Effects 0.000 description 5
- 230000004044 response Effects 0.000 description 5
- 241000282412 Homo Species 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 4
- 238000013138 pruning Methods 0.000 description 4
- 238000012552 review Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 235000019557 luminance Nutrition 0.000 description 3
- 230000000873 masking effect Effects 0.000 description 3
- 230000002829 reductive effect Effects 0.000 description 3
- 244000290333 Vanilla fragrans Species 0.000 description 2
- 235000009499 Vanilla fragrans Nutrition 0.000 description 2
- 235000012036 Vanilla tahitensis Nutrition 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 2
- 230000006835 compression Effects 0.000 description 2
- 238000007906 compression Methods 0.000 description 2
- 230000001419 dependent effect Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000000670 limiting effect Effects 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000012946 outsourcing Methods 0.000 description 2
- 238000013515 script Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 241001342895 Chorus Species 0.000 description 1
- 229920000742 Cotton Polymers 0.000 description 1
- HJBWJAPEBGSQPR-UHFFFAOYSA-N DMCA Natural products COC1=CC=C(C=CC(O)=O)C=C1OC HJBWJAPEBGSQPR-UHFFFAOYSA-N 0.000 description 1
- 241000287828 Gallus gallus Species 0.000 description 1
- 206010028916 Neologism Diseases 0.000 description 1
- 241000030560 Styx Species 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 230000004931 aggregating effect Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000009118 appropriate response Effects 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 239000003795 chemical substances by application Substances 0.000 description 1
- 235000019219 chocolate Nutrition 0.000 description 1
- 239000000470 constituent Substances 0.000 description 1
- HAORKNGNJCEJBX-UHFFFAOYSA-N cyprodinil Chemical compound N=1C(C)=CC(C2CC2)=NC=1NC1=CC=CC=C1 HAORKNGNJCEJBX-UHFFFAOYSA-N 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000003292 diminished effect Effects 0.000 description 1
- 238000005553 drilling Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 239000000446 fuel Substances 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- PCHJSUWPFVWCPO-UHFFFAOYSA-N gold Chemical compound [Au] PCHJSUWPFVWCPO-UHFFFAOYSA-N 0.000 description 1
- 239000010931 gold Substances 0.000 description 1
- 229910052737 gold Inorganic materials 0.000 description 1
- 244000144980 herd Species 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 229920001690 polydopamine Polymers 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000010079 rubber tapping Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- GOLXNESZZPUPJE-UHFFFAOYSA-N spiromesifen Chemical compound CC1=CC(C)=CC(C)=C1C(C(O1)=O)=C(OC(=O)CC(C)(C)C)C11CCCC1 GOLXNESZZPUPJE-UHFFFAOYSA-N 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000008685 targeting Effects 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 230000000699 topical effect Effects 0.000 description 1
- 210000003462 vein Anatomy 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/18—Legal services
- G06Q50/184—Intellectual property management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/10—Office automation; Time management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q20/00—Payment architectures, schemes or protocols
- G06Q20/08—Payment architectures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q20/00—Payment architectures, schemes or protocols
- G06Q20/22—Payment schemes or models
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/01—Social networking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/18—Legal services
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/23—Processing of content or additional data; Elementary server operations; Server middleware
- H04N21/234—Processing of video elementary streams, e.g. splicing of video streams or manipulating encoded video stream scene graphs
- H04N21/23418—Processing of video elementary streams, e.g. splicing of video streams or manipulating encoded video stream scene graphs involving operations for analysing video streams, e.g. detecting features or characteristics
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/25—Management operations performed by the server for facilitating the content distribution or administrating data related to end-users or client devices, e.g. end-user or client device authentication, learning user preferences for recommending movies
- H04N21/254—Management at additional data server, e.g. shopping server, rights management server
- H04N21/2541—Rights Management
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/25—Management operations performed by the server for facilitating the content distribution or administrating data related to end-users or client devices, e.g. end-user or client device authentication, learning user preferences for recommending movies
- H04N21/258—Client or end-user data management, e.g. managing client capabilities, user preferences or demographics, processing of multiple end-users preferences to derive collaborative data
- H04N21/25808—Management of client data
- H04N21/25841—Management of client data involving the geographical location of the client
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/25—Management operations performed by the server for facilitating the content distribution or administrating data related to end-users or client devices, e.g. end-user or client device authentication, learning user preferences for recommending movies
- H04N21/258—Client or end-user data management, e.g. managing client capabilities, user preferences or demographics, processing of multiple end-users preferences to derive collaborative data
- H04N21/25866—Management of end-user data
- H04N21/25891—Management of end-user data being end-user preferences
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/80—Generation or processing of content or additional data by content creator independently of the distribution process; Content per se
- H04N21/83—Generation or processing of protective or descriptive data associated with content; Content structuring
- H04N21/835—Generation of protective data, e.g. certificates
- H04N21/8352—Generation of protective data, e.g. certificates involving content or source identification data, e.g. Unique Material Identifier [UMID]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/80—Generation or processing of content or additional data by content creator independently of the distribution process; Content per se
- H04N21/83—Generation or processing of protective or descriptive data associated with content; Content structuring
- H04N21/835—Generation of protective data, e.g. certificates
- H04N21/8355—Generation of protective data, e.g. certificates involving usage data, e.g. number of copies or viewings allowed
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/80—Generation or processing of content or additional data by content creator independently of the distribution process; Content per se
- H04N21/83—Generation or processing of protective or descriptive data associated with content; Content structuring
- H04N21/835—Generation of protective data, e.g. certificates
- H04N21/8358—Generation of protective data, e.g. certificates involving watermark
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/80—Generation or processing of content or additional data by content creator independently of the distribution process; Content per se
- H04N21/83—Generation or processing of protective or descriptive data associated with content; Content structuring
- H04N21/84—Generation or processing of descriptive data, e.g. content descriptors
- H04N21/8405—Generation or processing of descriptive data, e.g. content descriptors represented by keywords
Definitions
- Web site operators sometimes receive DMCA “take down” notices from media companies, alleging that content hosted on their web sites is copyrighted and should not be distributed. There is a growing need for automated tools to help web site operators pro-actively identify such content and treat it in a manner that might avoid the need for take-down notices. This need perhaps is felt most acutely by so-called “social networking” sites, to which individual users upload audio, video and picture files—content that is sometimes original, and sometimes not (and sometimes a combination).
- Another approach is to try and identify the content by pattern recognition techniques (sometimes termed “fingerprinting” or “robust hashing”). Once the content is identified, a metadata database can be consulted to determine whether distribution of the content should be allowed or prohibited. (Such techniques are detailed, e.g., in Haitsma, et al, “A Highly Robust Audio Fingerprinting System,” Proc.
- Part of the difficulty is that some of the content uploaded to web sites may include copyrighted material, yet qualify as “fair use”—such as parody, or commentary/criticism.
- “Tolerated use” is a stepchild of fair use, and encompasses arguably infringing uses that are commonly overlooked by rights holders for reasons such as concern about adverse publicity, or out of desire for the exposure that such use affords.)
- Existing automated techniques do not make provision for “fair use” (nor for “tolerated use”). Instead, known techniques typically flag as objectionable any content that is determined to include any copyrighted material.
- Described below is an illustrative arrangement that allows a more nuanced assessment of content data—one that responds differently, depending on context, environmental factors, and/or other circumstances.
- FIG. 1 shows an illustrative system employing some of the technology detailed herein.
- FIG. 2 shows some of the factors that may be considered in determining treatment of content.
- FIG. 3 is a flow diagram for an illustrative process.
- uploaded content e.g., a video uploaded to MySpace, or YouTube
- identify portion(s) that have been derived from one or more pre-existing (reference) works e.g., a technique such as fingerprinting, or digital watermark decoding
- a technique such as fingerprinting, or digital watermark decoding
- the video may further include a government archive clip of Ronald Reagan, threatening to veto tax-increase legislation by using the same phrase, as well as a clip from the movie Back to the Future III, where Marty utters the same phrase.
- the video may also be found to include, as an underlying soundtrack, the entirety of the Eminem's song “Mosh.”
- Roboto by Styx (02.04-02.14); “Break Dance (Electric Boogie),” by West Street Mob (02.14-02.28); “Walk Like An Egyptian,” by The Bangles (02.28-02.36); “The Chicken Dance,” by Bob Kames (02.36-02.42); “Mony Mony,” by Billy Idol (02.42-02.57); “Ice Ice Baby,” by Vanilla Ice (02.57-03.11); “U Can't Touch This,” by MC Hammer (03.12-03.42); “Love Shack,” by The B-52's (03.42-03.46); “Apache,” by The Sugarhill Gang (03.46-04.02); “Jump Around,” by House of Pain (04.02-04.15); “Baby Got Back,” by Sir Mix-A-Lot (04.15-04.22); “Tubthumping,” by Chumbawamba (04.22-04.32); “What Is Love,” by Haddaway (04.32-04.40); “Cotton Eyed Joe,” by Rednex (04.4
- Video “mash-ups” such as these are increasingly common among user-generated content posted to social networking websites.
- An exemplary embodiment of the present technology considers a plurality of factors and/or parameters in making an automated decision. Such considerations can include (by way of example and not limitation):
- the length can be, e.g., in absolute terms (e.g., 10 seconds), or in percentage of the pre-existing work (e.g., 10 seconds may comprise 20% of a 50 second video, or a tiny fraction of a two hour movie).
- Different procedures may be invoked, depending on the length. For example, if the entirety of a pre-existing work is used, then free distribution may be prohibited (and a notice may be electronically sent to the submitter explaining same), or the system may permit free distribution but share 75% of associated ad revenue with the copyright holder of the work. If two seconds or less is used, the system may permit free distribution.
- the system may permit free distribution, but send a notice to the copyright holder alerting it to the use, and periodically alert the copyright holder to the content's popularity ranking on the site. If from five to ten seconds is used, and such portion comprises less than 10% of the entirety of a copyrighted work, the system may proceed as just-stated, with 10% of associated ad revenue shared with the copyright holder—unless copyright in the work is held by Company A, in which case no distribution is permitted (and a notification is sent to the submitter).
- the system may proceed as just-stated but with a 15% ad revenue share, unless copyright in the work is held by one of several specified Japanese rights-holders, in which case no distribution is permitted.
- any applicable ad-revenue share otherwise determined to be applicable may be increased by 25%. If the portion is taken from the chorus or title phrase of a song, any applicable ad-revenue share otherwise determined to be applicable may be increased by 33%.
- the system may determine that no restriction should arise by use of such clip—regardless of its length. If the portion is determined to have been derived from a copyrighted work owned by Company B, then the system may notify the Company if the content is sufficiently popular that it ranks among the 100 top videos downloaded from the site for any given day or week.
- the system may edit the content to remove that portion—if less than 5 seconds in length (possibly substituting a screen saying “Video omitted at request of Company C,” or another message established by Company C or the web site, if the copyrighted work is visual), or, if longer, prohibit distribution of the content.
- the portion is a video clip is determined to depict Individual D, who may enjoy a protectable right of publicity, then a notice may be dispatched to a last-known contact address for that person, and 10% of any associated ad revenue may be deposited into a reserve account to help fund possible claims against the web site operator.
- watermark- or fingerprint-indexed metadata permits identification of actors depicted in a video clip
- contract terms between the web site and the Screen Actors' Guild may specify, e.g., that each actor is entitled to 0.25 cents each time the content is downloaded.
- portion includes music in the licensing catalog of the Harry Fox agency, or ASCAP/BMI, a royalty payment of 0.1 cents for each ten seconds of music, up to a maximum of 0.4 cents, may be accrued for payment each time the content is downloaded.
- parties in addition to rights-holders may be considered in determining treatment of content.
- One such party is the person from whom the content is uploaded.
- Other potentially relevant parties include those in the chain of content distribution that led to the content finally being uploaded to the web site. (Such chain may be determined, e.g., by watermark data added by various parties involved in the content's distribution.)
- Ten percent of associated ad revenue may be provided to the person who uploaded the content, and a further five percent may be distributed among earlier parties in the chain of distribution.
- Metadata can be consulted to determine the language of the portion, or same can be determined by machine recognition techniques. If the portion is in Japanese, and the content is downloaded to a user in the United States, any sharing of ad revenue otherwise owing may be reduced by 50%. If a Japanese language clip is included in a video mash-up that includes no other Japanese language content (e.g., the rest of the mash-up comprises Spanish or English language content), then any sharing of ad revenue otherwise owing may be increased by 25%.
- a copyright levy may be paid to the royalty collecting society in the work's country of origin.
- the amount of this levy can be fixed (e.g., a penny per download), or it can be a share of advertising revenue (with different share percentages for content downloaded into different countries). These amounts can naturally also depend on others of these listed factors.
- Such metadata may specify, for example, that unlimited distribution is permitted, provided written attribution is provided to a specified rights-holder.
- the system can overlay a text legend on the video portion saying “Video used courtesy of Party E.” Such overlay can be accomplished on-the-fly, each time the content is downloaded. Or the content can be edited once to conform to the specified requirement, and stored in this edited fashion.
- the metadata may also specify terms of use with which the system can comply, e.g., this content can be used only if advertising is suppressed while the content is rendered.
- the metadata can also specify the amount and payee of required royalty payments. The system can make an automated assessment of such terms, and act accordingly.
- the system may apply a threshold cost limit (e.g., 1.5 cents) and permit redistribution of the content only if the required royalty payment is beneath this threshold.
- the system may further take other factors into account. For example, if the portion is recognized to be from one of several items of content that are known to be popular on the site (e.g., historical download data may show that all clips from The Daily Show by Jon Stewart fall into this category), it can triple the threshold cost limit otherwise applied.
- Metadata can be literally conveyed with content as noted above, it is often stored separately from the content, and indexed by an identifier derived from the content.
- the identifier may be derived by fingerprint technology (including, e.g., pattern matching), or by watermark decoding.
- the metadata may be stored in a database maintained by a party having rights to the content, or by a third party—including the web site operator.
- a fingerprint-based identifier derived from content may serve as a pointer to a database record maintained by Hollywood content producer A.
- the web site may copy that metadata and mirror it in a database kept locally with the site's web server.
- Digimarc operates commercial metadata services, ImageBridge and MyPictureMarc, which, when queried by watermark-decoded content identifiers, return associated metadata.
- Another metadata service is detailed in patent application Ser. No. 11/614,947 (now published as US20070208711).
- Other providers such as Flickr, del.icio.us, and Google Video, likewise serve as metadata repositories for content.
- metadata repositories can provide data relating to many of the other factors noted herein (e.g., identification of rights holders, language, terms of use, etc.).
- the metadata may associate a Creative Commons license with the work.
- the “Attribution Non-commercial”license is similar, but without the requirement that derivative works be licensed on the same terms.
- “Attribution Share Alike” is similar, but allowing commercial uses (and requiring licensing on the same basis).
- Other, more specialized, licenses are also available, e.g., “Sampling Licenses” for re-use of snippets, “Public Domain Dedication” to free works from all copyright protection, “Music Sharing Licenses” for fan use, “Developing Nations License” with less restrictive terms in low income countries, etc.
- Each of these licenses has abbreviated descriptors, such as “by-nc” for the “Attribution Non-commercial” license.
- a clip from a Beatles song will often warrant a different treatment than a clip from an undiscovered garage band.
- Popularity of a pre-existing work can be determined in numerous ways. One is by reference to industry “Top” lists, such as those published by Billboard magazine (and available on the Billboard web site), etc. Another is by reference to popularity indices maintained by content vendors such as Amazon (the Amazon Sales Rank) and Apple/iTunes. Once content is identified, the system can assess its popularity by reference to one or more such data sources. If the popularity of the work exceeds a threshold (e.g., a ranking in the top 100 of any of Billboard's lists, or an Amazon Sales Rank of 500 or above), then treatment of such a portion can be different than that afforded portions derived from less popular works. For example, a fee may be charged to users who download the content. Or a portion of ad-related revenue may be provided to the copyright holder(s) of the constituent portions, or added to a reserve account to fund possible claims.
- the uploaded content's own popularity may influence its manner of treatment. For example, once a video breaks into YouTube's 100 Most Viewed videos of the day or week, it may merit different treatment than video that is rarely viewed. As noted in the preceding paragraph, a fee may be charged, or a portion of ad-related revenue may be earmarked.
- the acts detailed herein are not invoked routinely, but are invoked only when the content breaks into a top echelon of popularity.
- a different per-download fee may be warranted if content is downloaded to Canada than Cuba.
- values can be associated with pre-existing works, e.g., by reference to the price of the work as distributed in CD/DVD form by Amazon. Beatles music is rarely discounted; grade B movies are often discounted. From such information, better-informed business decisions may be made about treatment of content.
- a relationship can be inferred, or established, between uploaded content and content earlier handled by the web site.
- a new episode of The Daily Show by Jon Stewart may be recognized, by metadata association, to be related to earlier episodes.
- Information about the earlier episodes e.g., how it was treated by the website; consumption data; etc.
- the evaluator may, e.g., determine whether a pre-existing work identified by fingerprinting is a true match, or a false negative.
- the evaluator may also classify the content by genre, e.g., home video, amateur music video to commercial sound track, etc. In some cases, this assessment can be referred out to non-employees of the web site, e.g., by using crowdsourcing techniques like Amazon's Mechanical Turk service.
- Content that borrows from dozens of pre-existing works may call for different treatment than content that is based on just a single pre-existing work. For example, any split of ad revenue will need to take into account the relatively large number of others works involved. If such a mash-up borrows exclusively from the catalog of one rights-holder, it might be regarded as helpful publicity for that catalog and therefor be entitled to move favorable usage terms than a mash-up that borrows from a variety of different catalogs.
- the analysis of the content—and sometimes circumstance and context—to determine treatment can be undertaken by the web site system at the time content is uploaded to the site (e.g., at “ingest”).
- the system can seek to examine the uploaded content to try and identify all possible pre-existing works from which portions may have been derived, or it may only check against a limited number of pre-existing works (e.g., those works owned by predetermined parties).
- the system can undertake these actions at a later time (e.g., after the content has demonstrated some user popularity—such as being downloaded 100 or 1000 times; or in response to a third party request—such as by a rights holder who notices an excerpt of copyrighted content hosted by the site).
- Certain of these acts can also be undertaken not by the web site operator, but by third parties (e.g., content owners), such as by automated web scripts that download content of interest, and assess same using the acts detailed above.
- a sample system 100 in which the present technology can be practiced is illustrated in FIG. 1 .
- a web site 10 includes a database 12 , a CPU 14 , a watermark decoder 16 , a fingerprint algorithm 18 , a memory 20 , and local rules 22 . (Not shown are the well known interfaces by which the web site receives uploaded content 24 from users, and dispatches downloaded content 26 to requesting users.)
- CPU 14 can comprise one or many processors, as is typical in web server architectures.
- Memory 20 can comprise a large RAM array in which data and instructions (e.g., operating system, applications software, etc.) are loaded for use by the CPU.
- Database 12 is an organized data structure in which various data is stored for later retrieval. Examples include audio and video uploaded by users, cached content metadata previously retrieved from other sources, and may include local rules 22 .
- Watermark decoder 16 examines content for steganographically-embedded (digitally watermarked) metadata, such as content identifiers or rights management data.
- Fingerprint algorithm 18 derives fingerprint data from content, to aid in its identification. (Decoder 16 and algorithm 18 may be implemented in hardware, but are more typically implemented using software loaded into memory 20 and executed on CPU 14 .)
- Local rules 22 can establish default behaviors of the system in different circumstances. Such rules may provide, for example, that: downloads exceeding 10 minutes should be made only to registered site users; shorter downloads can be made to unregistered users; no content should be downloaded to IP addresses associated with Country X; advertising—sourced from Google based on metadata keywords provided to Google by the site—should be displayed with the content; etc.
- the content 24 uploaded from users may include metadata.
- This metadata can be “in-band” (e.g., steganographically embedded in audio or video data), or “out-of-band” (e.g., conveyed in file title, header fields, etc). It may be literally included, or a code or index pointer may be conveyed—by which corresponding metadata from a data repository can be retrieved.
- the metadata can include rules by which use of the content should be governed (e.g., a descriptor, such as “by-nc,” which indicates that the terms of the Creative Commons “Attribution, Non-Commercial” license apply).
- the metadata can also include information that is used in applying rules (e.g., the identification of rights-holders, the language of the content, etc.).
- Content 26 downloaded to users can similarly include metadata.
- This metadata can be of various types, including usage rules, serialization data, credit card of purchaser, etc.
- the content is steganographically encoded with keywords determined through processes like those detailed herein. For example, a video of a football game may be encoded with the term (or an index that corresponds to the term) “football.”
- Such steganographically-encoded metadata can serve to trigger appropriate responses at the downloader's device, e.g., presenting an advertisement for Jeep products.
- the operation often starts by examining the content to identify same.
- watermarking steganography
- fingerprinting can be employed for this purpose.
- Watermarking is advantageous in that a watermarked work (or portion thereof) can be identified with a high degree of confidence.
- Fingerprinting compares characteristics of the content data against a database of reference data to try and make a best-guess as to the content's identity.
- the identification operates with a relatively high degree of granularity (e.g., on the order of seconds for temporal media) to identify all components of which the content may be comprised.
- Such identification techniques commonly make use of one or more accessory databases 28 , 30 to resolve a watermark or fingerprint identifier (e.g., A75C20) into a more generally useful content identifier.
- content identifier can comprise, e.g., a textual description of the content, a alphanumeric identifier such as ISAN (International Standard Audiovisual Number), etc.
- the database 28 / 30 can also provide other metadata, including language, copyright holder, actor names, terms of use, and applicable rules.
- the databases 28 , 30 can be general purpose in scope (such as the MusicID database service offered by Gracenote), or may encompass the content catalog of only one or a few companies (e.g., a Hollywood studio may maintain a database relating to its specific offerings).
- Many television broadcasts are digitally watermarked with metadata identifying the broadcasting station and the time of broadcast, and a corresponding database may be queried by system 100 with these data to learn the identity of the marked content. (See, e.g., Nielsen's U.S. Pat. Nos. 7,117,513 and 6,647,548, and publication US20060171474).
- System 100 may consult several such databases in an attempt to identify, and collect metadata, about content.
- rule data can be gleaned from the content itself (e.g., via metadata), or determined by reference to the web site's own rules database 22 , the system may also consult external sources of rules and related data, such as databases 32 a and 32 b .
- a studio may provide a database in which rule information for particular content (e.g., indexed by an ISAN identifier) is stored. This allows the content proprietors to more quickly adapt terms to changing market conditions (e.g., in order to promote an upcoming movie sequel, the studio may temporarily relax rules relating to the original movie).
- a metadata database 34 is shown in FIG. 1 .
- the depicted database may be one of many.
- Such databases can comprise, e.g., Flickr, Del.icio.us, CMDS (detailed in application US20070208711), etc., and can again provide both rules, and related data on which rule execution may be based.
- the system 100 may also make reference to rules and/or data maintained by a licensing agency 36 , such as BMI, ASCAP, Harry Fox.
- Data source 36 can also comprise a rights server associated with DRM systems, such as Microsoft's Rights Management Services, Windows Media Rights Manager, Windows Media DRM 10 , etc. (Applicable rules may require the system 100 to obtain licenses for distribution of certain content.)
- Metadata obtained from one data repository can be used to query further metadata sources (e.g., a database that lists actors and publication data for a movie in response to an input ISAN query).
- Data from that second repository can be used to find further data in third and additional data repositories—providing a large universe of metadata resources from which the system can draw.
- each of these repositories can includes rule information, which may be aggregated and combined to accurately determine appropriate treatment of the content, in accordance with the various factors in this disclosure.
- the system 100 can store data relating to same in database 12 for later reference—obviating the need to poll all the remote sources repeatedly. (Some rules may be based on information that changes over time; this information will periodically have to be updated.)
- the web site serves as a virtual hub—querying remote sources and having resulting information returned to the web site. Naturally, this is not critical.
- Other architectural arrangements (a few of which are indicated by dashed arrows) can be used—with information being exchanged between remote computers, without the web site always acting as intermediary.
- FIG. 2 is a conceptual view of one particular implementation, in which content is identified, and corresponding rules and related information are then gathered from a variety of sources and applied in determining how the content should be treated.
- the content treatment may also be a function of other data, such as the current date or time, and the identity or geographical location of the user who uploaded or requests download of the content.
- FIG. 3 shows a sample process flow.
- the content is first examined to determine if any part is identifiable as derived from a copyrighted work (e.g., by fingerprinting or watermark decoding). If no pre-existing work can be identified, then a rule allowing free distribution is invoked. If the content is a movie, and it is found to be an unaltered copy of the entirety of a copyrighted movie (i.e., with complete audio and video tracks), then a rule prohibiting free distribution is invoked. A similar rule is applied if the content is determined to be audio, and it is found to be an unaltered copy of the entirety of a copyrighted audio track. If neither of those conditions is met, the process applies rules that determine appropriate treatment, based on nature and extent of copying, and other data.
- a copyrighted work e.g., by fingerprinting or watermark decoding
- Another sample method examines uploaded content having both audio and video tracks. If the audio and video tracks are determined to be unaltered copies of respective audio and video portions of a pre-existing work, then the content is treated in a first manner (e.g., distribution is prohibited; or a fee is charged to downloading users, and 80% of the fee is paid to an associated rights-holder; etc). If the audio is determined to be an unaltered copy of the audio portion of a pre-existing work, but no such determination is made concerning the video, then perhaps the content is a home music video, with a person performing to a commercial music track.
- a first manner e.g., distribution is prohibited; or a fee is charged to downloading users, and 80% of the fee is paid to an associated rights-holder; etc.
- the content is treated in a second manner, e.g., referred to a human evaluator to determine if the work might be regarded as a parody, and distribution thus permitted. If the video is determined to be an unaltered copy of the video portion of a pre-existing work, but the audio is different, then it might be a commercially valuable video, dubbed into a foreign language. In such circumstance the content may be treated in a third manner, such as prohibiting distribution, or referred to a human evaluator for assessment.
- the content identity can be determined (or inferred) by other information, or by circumstance.
- an uploaded file that includes an identifying ISAN number in a header field, together with a digital signature evidencing that the file has not been altered since being released by the authoring Hollywood studio.
- a file may be retrieved from the site by a remote web crawler or Perl script, in response to title keywords entered in a “Search” dialog box; results that match the entered title keywords may be inferred to include the searched-for content.
- Bayesian techniques can be advantageously employed in help determine appropriate treatment for content objects and excerpts.
- content including, without limitation, network routers, firewalls, cell phones, set-top boxes, email servers and gateways, user computers, PDAs, blogging web sites, RSS servers and clients, indexing web crawlers and other web agents, peer-to-peer systems, video games, etc.
- the rules employed in the detailed arrangements can be structured, stored, distributed, and applied in different ways.
- rules specific to particular content items there may be rules specific to particular content creators, content owners, content hosts (e.g., YouTube vs. MySpace), etc.
- Web 2.0 technologies e.g., employing standardized descriptors (XML and other semantically-valid tags, etc.), Ajax- or Flex-based UIs, folksonomies (collaborative tagging), content syndication, etc.
- the Mechanical Turk service may be regarded as a structured implementation of a technology commonly termed “crowdsourcing”—employing a group of outsiders to perform a task. Wikipedia explains:
- Amazon's Mechanical Turk system or similar crowdsourcing arrangements, are employed to match a first item of visual or audio entertainment content to a counterpart in a universe of such items.
- a user social networking site such as YouTube (now Google) that distributes “user generated content” (e.g., video files), and employs fingerprinting to recognize media content that should not be distributed.
- the site may check a video file at the time of its uploading with a fingerprint recognition system (e.g., of the sort offered by Audible Magic, or Gracenote). If no clear match is identified, the video may be indexed and stored on YouTube's servers, available for public downloading. Meanwhile, the content can be queued for review by one or more crowdsource reviewers. They may recognize it as a clip from the old TV sitcom “I Love Lucy”—perhaps digitally rotated 3 degrees to avoid fingerprint detection. This tentative identification is returned to YouTube from the API call.
- a fingerprint recognition system e.g., of the sort offered by Audible Magic, or Gracenote
- YouTube can check the returning metadata against a title list of works that should not be distributed (e.g., per the request of copyright owners), and may discover that “I Love Lucy” clips should not be distributed. It can then remove the content from public distribution. Additionally, the fingerprint database can be updated with the fingerprint of the rotated version of the I Love Lucy clip, allowing it to be immediately recognized the next time it is encountered.
- a related technology is facial recognition—matching an unknown face to a reference database of facial images. Again, each of the faces is distilled down to a characteristic set of features, and a match is sought between an unknown feature set, and feature sets corresponding to reference images.
- the feature set may comprise eigenvectors or shape primitives.
- Patent documents particularly concerned with such technology include US20020031253, U.S. Pat. Nos. 6,292,575, 6,301,370, 6,430,306, 6,466,695, and 6,563,950.
- one approach is to prune the database—identifying excerpts thereof that are believed to be relatively likely to have a match, and limiting the search to those excerpts (or, similarly, identifying excerpts that are believed relatively unlikely to have a match, and not searching those excerpts).
- the database search may locate several reference fingerprints that are similar to the fingerprint of the unknown content.
- the identification process then seeks to identify a “best” match, using various algorithms.
- Such content identification systems can be improved by injecting a human into the process—by the Mechanical Turk service or similar systems.
- the content identification system makes an assessment of the results of its search, e.g., by a score.
- a score of 100 may correspond to a perfect match between the unknown fingerprint and a reference fingerprint.
- Lower scores may correspond to successively less correspondence. (At some lower score, S x , (perhaps 60) the system may decide that there is no suitable match, and a “no-match” result is returned, with no identification made.)
- S y (perhaps 70) the system may be sufficiently confident of the result that no human intervention is necessary.
- the system may make a call through the Mechanical Turk service for assistance.
- the Mechanical Turk can be presented the unknown content (or an excerpt thereof), and some reference content, and asked to make a comparison.
- the reference content may be stored in the fingerprint database, or may be readily obtainable through use of a link stored in the reference database.
- a single item of reference content can be provided for comparison with the unknown content, or several items of reference content can be provided. (Again, excerpts may be used instead of the complete content objects.
- the content might be processed before sending to the crowdsource engine, e.g., removing metadata (such as personally identifiable information: name, driver license number, etc.) that is printed on, or conveyed with, the file.)
- the requested comparison can take different forms.
- the service can be asked simply whether two items appear to match. Or it can be asked to identify the best of several possible matches (or indicate that none appears to match). Or it can be asked to give a relative match score (e.g., 0-100) between the unknown content and one or more items reference content.
- a relative match score e.g., 0-100
- a query is referred to several different humans (e.g., 2-50) through the Mechanical Turk service, and the returned results are examined for consensus on a particular answer.
- queries e.g., does Content A match Content B? Or is Content A a better match to Content C?
- a threshold of consensus e.g., 51%, 75%, 90%, 100%
- the scores returned from plural such calls may be combined to yield a net result.
- the high and/or low and/or outlier scores may be disregarded in computing the net result; weighting can sometimes be employed, as noted below.
- the data returned from the Mechanical Turk calls may serve as a biasing factor, e.g., pushing an algorithmically determined output one way or another, to yield a final answer (e.g., a net score).
- a final answer e.g., a net score
- the data returned from the Mechanical Turk calls may be treated as a definitive answer—with results from preceding processes disregarded.
- the database search may reveal several candidate matches, all with comparable scores (which may be above the threshold S y ). Again, one or more calls to the Mechanical Turk service may be invoked to decide which match is the best, from a subjective human standpoint.
- the Mechanical Turk service can be invoked even in situations where the original confidence score is below the threshold, S x , which is normally taken as indicating “no match.” Thus, the service can be employed to effectively reduce this threshold—continuing to search for potential matches when the rote database search does not yield any results that appear reliable.
- a database may be organized with several partitions (physical or logical), each containing information of a different class.
- the data may be segregated by subject gender (i.e., male facial portraits, female facial portraits), and/or by age (15-40, 30-65, 55 and higher—data may sometimes be indexed in two or more classifications), etc.
- the data may be segregated by topical classification (e.g., portrait, sports, news, landscape).
- topical classification e.g., portrait, sports, news, landscape
- the data may be segregated by type (spoken word, music, other).
- Each classification in turn, can be further segregated (e.g., “music” may be divided into classical, country, rock, other). And these can be further segregated (e.g., “rock” may be classified by genre, such as soft rock, hard rock, Southern rock; by artist, e.g., Beatles, Rolling Stones, etc.).
- a call to the Mechanical Turk can be made, passing the unknown content object (or an excerpt thereof) to a human reviewer, soliciting advice on classification.
- the human can indicate the apparent class to which the object belongs (e.g., is this a male or female face? Is this music classical, country, rock, or other?). Or, the human can indicate one or more classes to which the object does not belong.
- the system can focus the database search where a correct match—if any—is more likely to be found (or avoid searching in unproductive database excerpts). This focusing can be done at different times. In one scenario it is done after a rote search is completed, in which the search results yield matches below the desired confidence level of S y . If the database search space is thereafter restricted by application of human judgment, the search can be conducted again in the limited search space. A more thorough search can be undertaken in the indicated subset(s) of the database. Since a smaller excerpt is being searched, a looser criteria for a “match” might be employed, since the likelihood of false-positive matches is diminished. Thus, for example, the desired confidence level S y might be reduced from 70 to 65. Or the threshold S x at which “no match” is concluded, may be reduced from 60 to 55. Alternatively, the focusing can be done before any rote searching is attempted.
- the result of such a human-focused search may reveal one or more candidate matches.
- the Mechanical Turk service may be called a second time, to vet the candidate matches—in the manner discussed above. This is one of several cases in which it may be desirable to cascade Mechanical Turk calls—the subsequent calls benefiting from the former.
- the first Mechanical Turk call aids in pruning the database for subsequent search.
- the second call aids in assessing the results of that subsequent search.
- Mechanical Turk calls of the same sort can be cascaded.
- the Mechanical Turk first may be called to identify audio as music/speech/other.
- a second call may identify music (identified per the first call) as classical/country/rock/other.
- a third call may identify rock (identified per the second call) as Beatles/Rolling Stones/etc.
- iterative calling of a crowdsourcing service a subjective judgment can be made that would be very difficult to achieve otherwise.
- human reviewers are pre-qualified as knowledgeable in a specific domain (e.g., relatively expert in recognizing Beatles music). This qualification can be established by an online examination, which reviewers are invited to take to enable them to take on specific tasks (often at an increased rate of pay). Some queries may be routed only to individuals that are pre-qualified in a particular knowledge domain. In the cascaded example just given, for example, the third call might be routed to one or more users with demonstrated expertise with the Beatles (and, optionally, to one or more users with demonstrated expertise with the Rolling Stones, etc.). A positive identification of the unknown content as sounding like the Beatles would be given more relative weight if coming from a human qualified in this knowledge domain.
- Calls to the Mechanical Turk service may request the human to provide metadata relevant to any content reviewed. This can include supposed artist(s), genre, title, subject, date, etc.
- This information (which may be ancillary to a main request, or may comprise the entirety of the request) can be entered into a database. For example, it can be entered into a fingerprint database—in association with the content reviewed by the human.
- data gleaned from Mechanical Turk calls are entered into the database, and employed to enrich its data—and enrich information that can be later mined from the database. For example, if unknown content X has a fingerprint F x , and through the Mechanical Turk service it is determined that this content is a match to reference content Y, with fingerprint F y , then a corresponding notation can be added to the database, so that a later query on fingerprint F x (or close variants thereof) will indicate a match to content Y. (E.g., a lookup table initially indexed with a hash of the fingerprint F x will point to the database record for content Y.)
- Calls to outsourcing engines involve a time lag before results are returned.
- the calling system can generally cope, or be adapted to cope, with such lags.
- a user social networking site such as YouTube (now Google) that distributes “user generated content” (e.g., video files), and employs fingerprinting to recognize media content that should not be distributed.
- the site may check a video file at the time of its uploading with a fingerprint recognition system (e.g., of the sort offered by Audible Magic, or Gracenote). If no clear match is identified, the video may be indexed and stored on YouTube's servers, available for public downloading. Meanwhile, the content can be queued for review by one or more crowdsource reviewers. They may recognize it as a clip from the old TV sitcom “I Love Lucy”—perhaps digitally rotated 3 degrees to avoid fingerprint detection. This tentative identification is returned to YouTube from the API call.
- a fingerprint recognition system e.g., of the sort offered by Audible Magic, or Gracenote
- YouTube can check the returning metadata against a title list of works that should not be distributed (e.g., per the request of copyright owners), and may discover that “I Love Lucy” clips should not be distributed. It can then remove the content from public distribution. (This generally follows a double-check of the identification by a YouTube employee.) Additionally, the fingerprint database can be updated with the fingerprint of the rotated version of the I Love Lucy clip, allowing it to be immediately recognized the next time it is encountered.
- the delivery can be interrupted.
- An explanatory message can be provided to the user (e.g., a splash screen presented at the interruption point in the video).
- Rotating a video by a few degrees is one of several hacks that can defeat fingerprint identification. (It is axiomatic that introduction of any new content protection technology draws hacker scrutiny. Familiar examples include attacks against Macrovision protection for VHS tapes, and against CSS protection for packaged DVD discs.) If fingerprinting is employed in content protection applications, such as in social networking sites (as outlined above) or peer-to-peer networks, its vulnerability to attack will eventually be determined and exploited.
- a well-known fingerprinting algorithm operates by repeatedly analyzing the frequency content of a short excerpt of an audio track (e.g., 0.4 seconds). The method determines the relative energy of this excerpt within 33 narrow frequency bands that logarithmically span the range 300 Hz-2000 Hz. A corresponding 32-bit identifier is then generated from the resulting data.
- a frequency band corresponds to a data bit “1” if its energy level is larger than that of the band above, and a “0” if its energy level is lower.
- Such a 32 bit identifier is computed every hundredth of a second or so, for the immediately preceding 0.4 second excerpt of the audio track, resulting in a large number of “fingerprints.”
- This series of characteristic fingerprints can be stored in a database entry associated with the track, or only a subset may be stored (e.g., every fourth fingerprint).
- the same calculation process is repeated.
- the resulting set of data is then compared against data earlier stored in the database to try and identify a match.
- various strategies can be employed to speed the search over a brute-force search technique, which yields unacceptable search times.
- the just-described technique is designed for audio identification, a similar arrangement can be used for video.
- the algorithm can use average luminances of blocks into which the image is divided as the key perceptual features. Again, a fingerprint can be defined by determining whether the luminance in each block is larger or smaller than the luminance of the preceding block.
- the reader may be familiar with different loudness profiles selectable on car radios, e.g., jazz, Talk, Rock, etc. Each applies a different frequency equalization profile to the audio, e.g., making bass notes louder if the Rock setting is selected, and quieter if the Talk setting is selected, etc. The difference is often quite audible when switching between different settings.
- the listener is generally unaware of which loudness profile is being employed. That is, without the ability to switch between different profiles, the frequency equalization imposed by a particular loudness profile is typically not noticed by a listener.
- the different loudness profiles yield different fingerprints.
- the 300 Hz energy in a particular 0.4 second excerpt may be greater than the 318 Hz energy.
- the situation may be reversed. This change prompts a change in the leading bit of the fingerprint.
- Audio multiband compression a form of processing that is commonly employed by broadcasters to increase the apparent loudness of their signal (most especially commercials).
- Such tools operate by reducing the dynamic range of a soundtrack—increasing the loudness of quiet passages on a band-by-band basis, to thereby achieve a higher average signal level. Again, this processing of the audio changes its fingerprint, yet is generally not objectionable to the listeners.
- Some formal attacks are based on psychoacoustic masking. This is the phenomena by which, e.g., a loud sound at one instant (e.g., a drum beat) obscures a listener's ability to perceive a quieter sound at a later instant. Or the phenomena by which a loud sound at one frequency (e.g., 338 Hz) obscures a listener's ability to perceive a quieter sound at a nearby frequency (e.g., 358 Hz) at the same instant. Research in this field goes back decades. (Modern watermarking software employs psychoacoustic masking in an advantageous way, to help hide extra data in audio and video content.)
- the algorithm detailed above would generate a fingerprint of ⁇ 011 . . . ⁇ from this data (i.e., 69 is less than 71, so the first bit is ‘0’; 71 is greater than 70, so the second bit is ‘1’; 70 is greater than 68, so the third bit is ‘1’).
- the fingerprint is now ⁇ 101 . . . ⁇ .
- Two of the three illustrated fingerprint bits have been changed. Yet the change to the audio excerpt is essentially inaudible.
- the exemplary fingerprinting technique noted above (which is understood to be the basis for Gracenote's commercial implementation, MusicID, built from technology licensed from Philips) is not unique in being vulnerable to various attacks. All fingerprinting techniques (including the recently announced MediaHedge, as well as CopySense and RepliCheck) are similarly believed to have vulnerabilities that can be exploited by hackers. (A quandary for potential adopters is that susceptibility of different techniques to different attacks has not been a focus of academic attention.)
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Signal Processing (AREA)
- Multimedia (AREA)
- Entrepreneurship & Innovation (AREA)
- Accounting & Taxation (AREA)
- Computer Security & Cryptography (AREA)
- Technology Law (AREA)
- Primary Health Care (AREA)
- Health & Medical Sciences (AREA)
- Operations Research (AREA)
- General Health & Medical Sciences (AREA)
- Finance (AREA)
- Databases & Information Systems (AREA)
- Development Economics (AREA)
- Quality & Reliability (AREA)
- Computing Systems (AREA)
- Educational Administration (AREA)
- Game Theory and Decision Science (AREA)
- Data Mining & Analysis (AREA)
- Computer Graphics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
Abstract
Content uploaded to a video distribution service is analyzed to determine if a portion matches part of any reference work. If a match is found, treatment of the uploaded content is determined based on stored rule data, and based on one or more factors. These factors can include, e.g., the length of the matching portion, the part of the reference work from which the matching portion was apparently copied, the confidence of the match, the popularity of the uploaded content (or of the reference work), the geography from which the content was uploaded (or the geography to which it is to be downloaded), etc. Further determinations may be made by human evaluators (including crowd-source approaches). A great variety of other features and arrangements are also detailed.
Description
- This application is a continuation of application Ser. No. 13/686,541, filed Nov. 27, 2012 (U.S. Pat. No. 10,242,415) which is a continuation-in-part of application Ser. No. 12/048,072, filed Mar. 13, 2008 (U.S. Pat. No. 9,179,200), and Ser. No. 12/114,612, filed May 2, 2008 (U.S. Pat. No. 8,341,412). Application Ser. No. 12/048,072 claims priority benefit from provisional application 60/894,859, filed Mar. 14, 2007. Application Ser. No. 12/114,612 application is a division of copending application Ser. No. 11/613,891, filed Dec. 20, 2006 (published as US20070162761).
- The following specification is essentially that of application Ser. No. 12/048,072. Appended to the end of this specification, as an appendix, is the text of application Ser. No. 12/114,612. This latter application was originally incorporated-by-reference into the Ser. No. 12/048,072 application (via a reference to its parent publication US20070162761).
- Web site operators sometimes receive DMCA “take down” notices from media companies, alleging that content hosted on their web sites is copyrighted and should not be distributed. There is a growing need for automated tools to help web site operators pro-actively identify such content and treat it in a manner that might avoid the need for take-down notices. This need perhaps is felt most acutely by so-called “social networking” sites, to which individual users upload audio, video and picture files—content that is sometimes original, and sometimes not (and sometimes a combination).
- Various techniques can be employed to automatically identify copyrighted content. One is to examine content data for a digital watermark embedded by the content owner to signal that the content is copyrighted and should not be reproduced. Such techniques are detailed, for example, in commonly-owned application Ser. No. 09/620,019, filed Jul. 20, 2000 (now U.S. Pat. No. 7,689,532), and patent publication US20020052885.
- Another approach is to try and identify the content by pattern recognition techniques (sometimes termed “fingerprinting” or “robust hashing”). Once the content is identified, a metadata database can be consulted to determine whether distribution of the content should be allowed or prohibited. (Such techniques are detailed, e.g., in Haitsma, et al, “A Highly Robust Audio Fingerprinting System,” Proc. Intl Conf on Music Information Retrieval, 2002; Cano et al, “A Review of Audio Fingerprinting,” Journal of VLSI Signal Processing, 41, 271, 272, 2005; Kalker et al, “Robust Identification of Audio Using Watermarking and Fingerprinting,” in Multimedia Security Handbook, CRC Press, 2005, and in patent documents WO02/065782, US20060075237, US20050259819, US20050141707, and US20020028000.)
- Other techniques and systems related to the technology detailed herein are disclosed in patent publications US20080051029, US20080059211, US20080027931, US20070253594, US20070242880, US20070220575, US20070208711, US20070175998, US20070162761, US20060240862, US20040243567, US20030021441, U.S. Pat. Nos. 7,185,201, 7,298,864 and 7,302,574, and in provisional application 61/016,321, filed Dec. 21, 2007.
- Part of the difficulty is that some of the content uploaded to web sites may include copyrighted material, yet qualify as “fair use”—such as parody, or commentary/criticism. (“Tolerated use” is a stepchild of fair use, and encompasses arguably infringing uses that are commonly overlooked by rights holders for reasons such as concern about adverse publicity, or out of desire for the exposure that such use affords.) Existing automated techniques do not make provision for “fair use” (nor for “tolerated use”). Instead, known techniques typically flag as objectionable any content that is determined to include any copyrighted material.
- Described below is an illustrative arrangement that allows a more nuanced assessment of content data—one that responds differently, depending on context, environmental factors, and/or other circumstances.
-
FIG. 1 shows an illustrative system employing some of the technology detailed herein. -
FIG. 2 shows some of the factors that may be considered in determining treatment of content. -
FIG. 3 is a flow diagram for an illustrative process. - Consider an embodiment that examines uploaded content (e.g., a video uploaded to MySpace, or YouTube), to identify portion(s) that have been derived from one or more pre-existing (reference) works. Using a technique such as fingerprinting, or digital watermark decoding, such a process may automatically determine that an uploaded video includes a ten second clip taken from the opening scene of the movie Dirty Harry, showing Clint Eastwood say, “Go ahead, make my day.” The video may further include a government archive clip of Ronald Reagan, threatening to veto tax-increase legislation by using the same phrase, as well as a clip from the movie Back to the Future III, where Marty utters the same phrase. The video may also be found to include, as an underlying soundtrack, the entirety of the Eminem's song “Mosh.”
- Another example (long popular on YouTube) is the “Evolution of Dance” video, which shows Judson Laipply dancing briefly to excerpts of 30 popular songs (parenthetical times are where the song clip appears in the six minute video): “Hound Dog,” by Elvis Presley (00.00-00.14); “The Twist,” by Chubby Checker (00.14-00.31); “Stayin' Alive,” by The Bee Gees (00.31-00.38); “Y.M.C.A.,” by The Village People (00.38-00.56); “Kung Fu Fighting,” by Carl Douglas (00.56-01.03); “Keep on Groovin',” by The Brady Bunch (01.03-01.17); “Greased Lightnin',” by John Travolta (01.17-01.28); “You Shook Me All Night Long,” by AC/DC (01.28-01.42); “Billie Jean,” by Michael Jackson (01.42-01.50); “Thriller,” by Michael Jackson (01.50-01.58); “Oompa Loompa,” by Willy Wonka, a soundtrack of the film: Willy Wonka & the Chocolate Factory (01.58-02.04); “Mr. Roboto,” by Styx (02.04-02.14); “Break Dance (Electric Boogie),” by West Street Mob (02.14-02.28); “Walk Like An Egyptian,” by The Bangles (02.28-02.36); “The Chicken Dance,” by Bob Kames (02.36-02.42); “Mony Mony,” by Billy Idol (02.42-02.57); “Ice Ice Baby,” by Vanilla Ice (02.57-03.11); “U Can't Touch This,” by MC Hammer (03.12-03.42); “Love Shack,” by The B-52's (03.42-03.46); “Apache,” by The Sugarhill Gang (03.46-04.02); “Jump Around,” by House of Pain (04.02-04.15); “Baby Got Back,” by Sir Mix-A-Lot (04.15-04.22); “Tubthumping,” by Chumbawamba (04.22-04.32); “What Is Love,” by Haddaway (04.32-04.40); “Cotton Eyed Joe,” by Rednex (04.40-05.01); “Macarena,” by Los Del Rio (05.01-05.06); “Bye Bye Bye,” by N'Sync (05.06-05.29); “Lose Yourself,” by Eminem (05.29-05.33); “Hey Ya!,” by Outkast (05.33-05.39); “Dirt Off Your Shoulder,” by Jay-Z (05.39-05.49); “Ice Ice Baby,” by Vanilla Ice (Lyrics played: “Yo, let's get outta here. Word to your mother”.) (05.49-05.52); “Bye Bye Bye,” by N'Sync (Lyrics played: “Bye, bye, bye”.) (05.52-06.00).
- Video “mash-ups” such as these are increasingly common among user-generated content posted to social networking websites.
- How should a web site handle such content? Should it allow or forbid free distribution? Should it allow distribution with conditions, such as with some recompense to relevant copyright holders? Etc.
- An exemplary embodiment of the present technology considers a plurality of factors and/or parameters in making an automated decision. Such considerations can include (by way of example and not limitation):
- 1. The Length of the Derived Portion.
- The length can be, e.g., in absolute terms (e.g., 10 seconds), or in percentage of the pre-existing work (e.g., 10 seconds may comprise 20% of a 50 second video, or a tiny fraction of a two hour movie). Different procedures may be invoked, depending on the length. For example, if the entirety of a pre-existing work is used, then free distribution may be prohibited (and a notice may be electronically sent to the submitter explaining same), or the system may permit free distribution but share 75% of associated ad revenue with the copyright holder of the work. If two seconds or less is used, the system may permit free distribution. If between two and five seconds is used, the system may permit free distribution, but send a notice to the copyright holder alerting it to the use, and periodically alert the copyright holder to the content's popularity ranking on the site. If from five to ten seconds is used, and such portion comprises less than 10% of the entirety of a copyrighted work, the system may proceed as just-stated, with 10% of associated ad revenue shared with the copyright holder—unless copyright in the work is held by Company A, in which case no distribution is permitted (and a notification is sent to the submitter). If between ten and fifteen seconds is used, the system may proceed as just-stated but with a 15% ad revenue share, unless copyright in the work is held by one of several specified Japanese rights-holders, in which case no distribution is permitted. Etc. Etc.
- 2. The Part of the Pre-Existing Content from which the Portion is Derived.
- For example, if the portion is taken from the first or last ten seconds, or from the first or last one percent of the content—whichever is smaller, any applicable ad-revenue share otherwise determined to be applicable may be increased by 25%. If the portion is taken from the chorus or title phrase of a song, any applicable ad-revenue share otherwise determined to be applicable may be increased by 33%.
- 3. An Identity of a Party Having Rights to the Pre-Existing Work.
- If the portion is determined to have been authored by an agency of the US government (e.g., a government movie clip of President Reagan), then the system may determine that no restriction should arise by use of such clip—regardless of its length. If the portion is determined to have been derived from a copyrighted work owned by Company B, then the system may notify the Company if the content is sufficiently popular that it ranks among the 100 top videos downloaded from the site for any given day or week. If the portion is determined to have been derived from a copyrighted work owned by Company C, then the system may edit the content to remove that portion—if less than 5 seconds in length (possibly substituting a screen saying “Video omitted at request of Company C,” or another message established by Company C or the web site, if the copyrighted work is visual), or, if longer, prohibit distribution of the content. If the portion is a video clip is determined to depict Individual D, who may enjoy a protectable right of publicity, then a notice may be dispatched to a last-known contact address for that person, and 10% of any associated ad revenue may be deposited into a reserve account to help fund possible claims against the web site operator. If watermark- or fingerprint-indexed metadata permits identification of actors depicted in a video clip, contract terms between the web site and the Screen Actors' Guild may specify, e.g., that each actor is entitled to 0.25 cents each time the content is downloaded. If the portion includes music in the licensing catalog of the Harry Fox agency, or ASCAP/BMI, a royalty payment of 0.1 cents for each ten seconds of music, up to a maximum of 0.4 cents, may be accrued for payment each time the content is downloaded.
- 4. An Identity of Another Party Affiliated with the Pre-Existing Work.
- Sometimes parties in addition to rights-holders may be considered in determining treatment of content. One such party is the person from whom the content is uploaded. Other potentially relevant parties include those in the chain of content distribution that led to the content finally being uploaded to the web site. (Such chain may be determined, e.g., by watermark data added by various parties involved in the content's distribution.) Ten percent of associated ad revenue may be provided to the person who uploaded the content, and a further five percent may be distributed among earlier parties in the chain of distribution.
- 5. a Language of the Pre-Existing Work, or Other Language(s) Found in the Content.
- Metadata can be consulted to determine the language of the portion, or same can be determined by machine recognition techniques. If the portion is in Japanese, and the content is downloaded to a user in the United States, any sharing of ad revenue otherwise owing may be reduced by 50%. If a Japanese language clip is included in a video mash-up that includes no other Japanese language content (e.g., the rest of the mash-up comprises Spanish or English language content), then any sharing of ad revenue otherwise owing may be increased by 25%.
- 6. a Country of Origin of the Pre-Existing Work.
- If the portion is determined to have been copied from a television sitcom produced in the European Union, a copyright levy may be paid to the royalty collecting society in the work's country of origin. The amount of this levy can be fixed (e.g., a penny per download), or it can be a share of advertising revenue (with different share percentages for content downloaded into different countries). These amounts can naturally also depend on others of these listed factors.
- 7. Metadata Conveyed with the Portion, e.g., by Watermark Data, Header Data, Etc.
- Such metadata may specify, for example, that unlimited distribution is permitted, provided written attribution is provided to a specified rights-holder. In such case, the system can overlay a text legend on the video portion saying “Video used courtesy of Party E.” Such overlay can be accomplished on-the-fly, each time the content is downloaded. Or the content can be edited once to conform to the specified requirement, and stored in this edited fashion. The metadata may also specify terms of use with which the system can comply, e.g., this content can be used only if advertising is suppressed while the content is rendered. The metadata can also specify the amount and payee of required royalty payments. The system can make an automated assessment of such terms, and act accordingly. For example, the system may apply a threshold cost limit (e.g., 1.5 cents) and permit redistribution of the content only if the required royalty payment is beneath this threshold. The system may further take other factors into account. For example, if the portion is recognized to be from one of several items of content that are known to be popular on the site (e.g., historical download data may show that all clips from The Daily Show by Jon Stewart fall into this category), it can triple the threshold cost limit otherwise applied.
- 8. Metadata Conveyed Apart from the Portion, e.g., in a Database at the Web Site or at a Remote Location.
- While metadata can be literally conveyed with content as noted above, it is often stored separately from the content, and indexed by an identifier derived from the content. The identifier may be derived by fingerprint technology (including, e.g., pattern matching), or by watermark decoding. The metadata may be stored in a database maintained by a party having rights to the content, or by a third party—including the web site operator. For example, a fingerprint-based identifier derived from content may serve as a pointer to a database record maintained by Hollywood content producer A. The web site may copy that metadata and mirror it in a database kept locally with the site's web server. Digimarc operates commercial metadata services, ImageBridge and MyPictureMarc, which, when queried by watermark-decoded content identifiers, return associated metadata. Another metadata service is detailed in patent application Ser. No. 11/614,947 (now published as US20070208711). Other providers, such as Flickr, del.icio.us, and Google Video, likewise serve as metadata repositories for content. In addition to the use cases just-given (in connection with metadata conveyed with the portion), such metadata repositories can provide data relating to many of the other factors noted herein (e.g., identification of rights holders, language, terms of use, etc.).
- (The metadata (e.g., from Flickr) may associate a Creative Commons license with the work. There are six main types of Creative Commons licenses. The most restrictive (while still allowing redistribution) is “Attribution Non-commercial No Derivatives,” which permits others to share the work—unchanged and non-commercially, provided the user gives the creator credit and a link-back. Next is “Attribution Non-commercial Share Alike,” which allows others to make derivative works non-commercially, provided they credit the creator, and such works licensed under the same terms. The “Attribution Non-commercial”license is similar, but without the requirement that derivative works be licensed on the same terms. “Attribution Share Alike” is similar, but allowing commercial uses (and requiring licensing on the same basis). “Attribution No Derivatives,” forbids making derivative works, but allows redistribution in original form, with credit to the creator. Least restrictive is “Attribution (by),” which allows others to make derivative works and use commercially, as long as they credit the creator. Other, more specialized, licenses are also available, e.g., “Sampling Licenses” for re-use of snippets, “Public Domain Dedication” to free works from all copyright protection, “Music Sharing Licenses” for fan use, “Developing Nations License” with less restrictive terms in low income countries, etc. Each of these licenses has abbreviated descriptors, such as “by-nc” for the “Attribution Non-commercial” license. These and other terms-of-use type metadata can be used to govern automatically the manner in which content is treated by a web site.)
- 9. a Popularity Metric Associated with the Pre-Existing Work (or the Uploaded Content).
- A clip from a Beatles song will often warrant a different treatment than a clip from an undiscovered garage band. Popularity of a pre-existing work can be determined in numerous ways. One is by reference to industry “Top” lists, such as those published by Billboard magazine (and available on the Billboard web site), etc. Another is by reference to popularity indices maintained by content vendors such as Amazon (the Amazon Sales Rank) and Apple/iTunes. Once content is identified, the system can assess its popularity by reference to one or more such data sources. If the popularity of the work exceeds a threshold (e.g., a ranking in the top 100 of any of Billboard's lists, or an Amazon Sales Rank of 500 or above), then treatment of such a portion can be different than that afforded portions derived from less popular works. For example, a fee may be charged to users who download the content. Or a portion of ad-related revenue may be provided to the copyright holder(s) of the constituent portions, or added to a reserve account to fund possible claims.
- Similarly, the uploaded content's own popularity may influence its manner of treatment. For example, once a video breaks into YouTube's 100 Most Viewed videos of the day or week, it may merit different treatment than video that is rarely viewed. As noted in the preceding paragraph, a fee may be charged, or a portion of ad-related revenue may be earmarked. In one embodiment, the acts detailed herein are not invoked routinely, but are invoked only when the content breaks into a top echelon of popularity.
- 10. A Jurisdiction from which the Web Site is Hosted, or from which the Content was Uploaded, or to which the Content is to be Distributed.
- A different per-download fee may be warranted if content is downloaded to Canada than Cuba.
- Similarly, different treatment may be appropriate based on the location from which the content originated, or even the physical location of the web site servers.
- 11. A Value Associated with the Pre-Existing Work.
- Sometimes values can be associated with pre-existing works, e.g., by reference to the price of the work as distributed in CD/DVD form by Amazon. Beatles music is rarely discounted; grade B movies are often discounted. From such information, better-informed business decisions may be made about treatment of content.
- 12. Historical Information Concerning Related Content.
- Sometimes a relationship can be inferred, or established, between uploaded content and content earlier handled by the web site. A new episode of The Daily Show by Jon Stewart may be recognized, by metadata association, to be related to earlier episodes. Information about the earlier episodes (e.g., how it was treated by the website; consumption data; etc.) can be retrieved and used to inform how the new episode should best be handled.
- 13. Data Relating to an Assessment of Said Content Provided by a Human Evaluator.
- Sometimes part or all of the content may be referred to a human evaluator for assessment. The evaluator may, e.g., determine whether a pre-existing work identified by fingerprinting is a true match, or a false negative. The evaluator may also classify the content by genre, e.g., home video, amateur music video to commercial sound track, etc. In some cases, this assessment can be referred out to non-employees of the web site, e.g., by using crowdsourcing techniques like Amazon's Mechanical Turk service. (Such arrangements are detailed, e.g., in application US20070162761.) Such techniques can be employed to determine whether use of a pre-existing work has indicia of fair use (considering factors such as the purpose and character of the use, the nature of the copyrighted work, the amount and substantiality of the portion copied, and the effect of the use upon the potential market).
- 14. Factors, Like Those Above, Relating to One or More Other Portions of the Content, which have been Derived from Other Pre-Existing Works.
- Content that borrows from dozens of pre-existing works (like the “Evolution of Dance” video noted above) may call for different treatment than content that is based on just a single pre-existing work. For example, any split of ad revenue will need to take into account the relatively large number of others works involved. If such a mash-up borrows exclusively from the catalog of one rights-holder, it might be regarded as helpful publicity for that catalog and therefor be entitled to move favorable usage terms than a mash-up that borrows from a variety of different catalogs.
- It is expressly contemplated that systems incorporating the presently-described technology may employ rules that are dependent on more than one of the factors noted above. In some cases, some factors may be dependent on others. All such permutations and combinations are expressly encompassed by this specification.
- The analysis of the content—and sometimes circumstance and context—to determine treatment can be undertaken by the web site system at the time content is uploaded to the site (e.g., at “ingest”). The system can seek to examine the uploaded content to try and identify all possible pre-existing works from which portions may have been derived, or it may only check against a limited number of pre-existing works (e.g., those works owned by predetermined parties). Alternatively, instead of performing these acts at ingest, the system can undertake these actions at a later time (e.g., after the content has demonstrated some user popularity—such as being downloaded 100 or 1000 times; or in response to a third party request—such as by a rights holder who notices an excerpt of copyrighted content hosted by the site). Certain of these acts can also be undertaken not by the web site operator, but by third parties (e.g., content owners), such as by automated web scripts that download content of interest, and assess same using the acts detailed above.
- A
sample system 100 in which the present technology can be practiced is illustrated inFIG. 1 . Aweb site 10 includes adatabase 12, aCPU 14, awatermark decoder 16, afingerprint algorithm 18, amemory 20, andlocal rules 22. (Not shown are the well known interfaces by which the web site receives uploadedcontent 24 from users, and dispatches downloadedcontent 26 to requesting users.) -
CPU 14 can comprise one or many processors, as is typical in web server architectures.Memory 20 can comprise a large RAM array in which data and instructions (e.g., operating system, applications software, etc.) are loaded for use by the CPU.Database 12 is an organized data structure in which various data is stored for later retrieval. Examples include audio and video uploaded by users, cached content metadata previously retrieved from other sources, and may includelocal rules 22.Watermark decoder 16 examines content for steganographically-embedded (digitally watermarked) metadata, such as content identifiers or rights management data.Fingerprint algorithm 18 derives fingerprint data from content, to aid in its identification. (Decoder 16 andalgorithm 18 may be implemented in hardware, but are more typically implemented using software loaded intomemory 20 and executed onCPU 14.) -
Local rules 22 can establish default behaviors of the system in different circumstances. Such rules may provide, for example, that: downloads exceeding 10 minutes should be made only to registered site users; shorter downloads can be made to unregistered users; no content should be downloaded to IP addresses associated with Country X; advertising—sourced from Google based on metadata keywords provided to Google by the site—should be displayed with the content; etc. - The
content 24 uploaded from users (e.g., audio, video, games, imagery, etc.) may include metadata. This metadata can be “in-band” (e.g., steganographically embedded in audio or video data), or “out-of-band” (e.g., conveyed in file title, header fields, etc). It may be literally included, or a code or index pointer may be conveyed—by which corresponding metadata from a data repository can be retrieved. As noted, the metadata can include rules by which use of the content should be governed (e.g., a descriptor, such as “by-nc,” which indicates that the terms of the Creative Commons “Attribution, Non-Commercial” license apply). The metadata can also include information that is used in applying rules (e.g., the identification of rights-holders, the language of the content, etc.). - (
Content 26 downloaded to users can similarly include metadata. This metadata can be of various types, including usage rules, serialization data, credit card of purchaser, etc. In one particular arrangement, the content is steganographically encoded with keywords determined through processes like those detailed herein. For example, a video of a football game may be encoded with the term (or an index that corresponds to the term) “football.” Such steganographically-encoded metadata can serve to trigger appropriate responses at the downloader's device, e.g., presenting an advertisement for Jeep products.) - When
content 24 is processed (bysystem 100, or otherwise), the operation often starts by examining the content to identify same. As noted, both watermarking (steganography) and fingerprinting can be employed for this purpose. Watermarking is advantageous in that a watermarked work (or portion thereof) can be identified with a high degree of confidence. Fingerprinting, in contrast, compares characteristics of the content data against a database of reference data to try and make a best-guess as to the content's identity. Desirably, the identification operates with a relatively high degree of granularity (e.g., on the order of seconds for temporal media) to identify all components of which the content may be comprised. - Such identification techniques commonly make use of one or more
accessory databases database 28/30 can also provide other metadata, including language, copyright holder, actor names, terms of use, and applicable rules. - The
databases system 100 with these data to learn the identity of the marked content. (See, e.g., Nielsen's U.S. Pat. Nos. 7,117,513 and 6,647,548, and publication US20060171474).System 100 may consult several such databases in an attempt to identify, and collect metadata, about content. - While rule data can be gleaned from the content itself (e.g., via metadata), or determined by reference to the web site's
own rules database 22, the system may also consult external sources of rules and related data, such asdatabases - Also shown in
FIG. 1 is ametadata database 34. Again, the depicted database may be one of many. Such databases can comprise, e.g., Flickr, Del.icio.us, CMDS (detailed in application US20070208711), etc., and can again provide both rules, and related data on which rule execution may be based. - The
system 100 may also make reference to rules and/or data maintained by alicensing agency 36, such as BMI, ASCAP, Harry Fox.Data source 36 can also comprise a rights server associated with DRM systems, such as Microsoft's Rights Management Services, Windows Media Rights Manager,Windows Media DRM 10, etc. (Applicable rules may require thesystem 100 to obtain licenses for distribution of certain content.) - Metadata obtained from one data repository (e.g., an ISAN identifier obtained from a database in response to an index decoded from an audio watermark) can be used to query further metadata sources (e.g., a database that lists actors and publication data for a movie in response to an input ISAN query). Data from that second repository, in turn, can be used to find further data in third and additional data repositories—providing a large universe of metadata resources from which the system can draw. (Conversion of format, protocol, classification, and/or other parameters may be required when using the results from one database to query another; such arrangements are detailed, e.g., in publication US20040243567.) The metadata in each of these repositories can includes rule information, which may be aggregated and combined to accurately determine appropriate treatment of the content, in accordance with the various factors in this disclosure.
- Once rules applicable to
content 24 are determined, thesystem 100 can store data relating to same indatabase 12 for later reference—obviating the need to poll all the remote sources repeatedly. (Some rules may be based on information that changes over time; this information will periodically have to be updated.) In the depictedsystem 100, the web site serves as a virtual hub—querying remote sources and having resulting information returned to the web site. Naturally, this is not critical. Other architectural arrangements (a few of which are indicated by dashed arrows) can be used—with information being exchanged between remote computers, without the web site always acting as intermediary. -
FIG. 2 is a conceptual view of one particular implementation, in which content is identified, and corresponding rules and related information are then gathered from a variety of sources and applied in determining how the content should be treated. (The content treatment may also be a function of other data, such as the current date or time, and the identity or geographical location of the user who uploaded or requests download of the content.) -
FIG. 3 shows a sample process flow. In this arrangement, the content is first examined to determine if any part is identifiable as derived from a copyrighted work (e.g., by fingerprinting or watermark decoding). If no pre-existing work can be identified, then a rule allowing free distribution is invoked. If the content is a movie, and it is found to be an unaltered copy of the entirety of a copyrighted movie (i.e., with complete audio and video tracks), then a rule prohibiting free distribution is invoked. A similar rule is applied if the content is determined to be audio, and it is found to be an unaltered copy of the entirety of a copyrighted audio track. If neither of those conditions is met, the process applies rules that determine appropriate treatment, based on nature and extent of copying, and other data. - Another sample method examines uploaded content having both audio and video tracks. If the audio and video tracks are determined to be unaltered copies of respective audio and video portions of a pre-existing work, then the content is treated in a first manner (e.g., distribution is prohibited; or a fee is charged to downloading users, and 80% of the fee is paid to an associated rights-holder; etc). If the audio is determined to be an unaltered copy of the audio portion of a pre-existing work, but no such determination is made concerning the video, then perhaps the content is a home music video, with a person performing to a commercial music track. In this case the content is treated in a second manner, e.g., referred to a human evaluator to determine if the work might be regarded as a parody, and distribution thus permitted. If the video is determined to be an unaltered copy of the video portion of a pre-existing work, but the audio is different, then it might be a commercially valuable video, dubbed into a foreign language. In such circumstance the content may be treated in a third manner, such as prohibiting distribution, or referred to a human evaluator for assessment.
- Having described and illustrated the principles of the technology with reference to specific implementations, it will be recognized that the technology can be implemented in many other, different, forms.
- For example, while the arrangements described above employ fingerprinting or watermark decoding to determine content identity, this is not always needed. In some arrangements, for example, the content identity can be determined (or inferred) by other information, or by circumstance. Consider an uploaded file that includes an identifying ISAN number in a header field, together with a digital signature evidencing that the file has not been altered since being released by the authoring Hollywood studio. Or a file may be retrieved from the site by a remote web crawler or Perl script, in response to title keywords entered in a “Search” dialog box; results that match the entered title keywords may be inferred to include the searched-for content.
- In some systems and methods, Bayesian techniques can be advantageously employed in help determine appropriate treatment for content objects and excerpts.
- While the detailed arrangements particularly focused on treatment of content ingested by “user generated content” sites such as YouTube, MySpace and Flickr, the same principles are likewise applicable by any other system that receives or processes content (including, without limitation, network routers, firewalls, cell phones, set-top boxes, email servers and gateways, user computers, PDAs, blogging web sites, RSS servers and clients, indexing web crawlers and other web agents, peer-to-peer systems, video games, etc.).
- It will be recognized that software instructions for programming a computer or processor in accordance with the teachings herein—as well as related data—can be stored on a computer-readable storage medium.
- The rules employed in the detailed arrangements can be structured, stored, distributed, and applied in different ways. In addition to rules specific to particular content items, there may be rules specific to particular content creators, content owners, content hosts (e.g., YouTube vs. MySpace), etc.
- Many implementations will make use of Web 2.0 technologies, e.g., employing standardized descriptors (XML and other semantically-valid tags, etc.), Ajax- or Flex-based UIs, folksonomies (collaborative tagging), content syndication, etc.
- To provide a comprehensive disclosure without unduly lengthening this specification, applicant incorporates by reference each of the documents mentioned herein (for all of their respective teachings—not just the feature(s) mentioned).
- The particular combinations of elements and features in the detailed embodiments are exemplary only; the interchanging, combination, and substitution of these teachings with other teachings in this and the incorporated-by-reference documents are expressly contemplated.
- Certain implementations of the technology make use of Amazon's Mechanical Turk service. Amazon's Mechanical Turk web site explains:
-
- Amazon Mechanical Turk provides a web services API for computers to integrate Artificial Artificial Intelligence directly into their processing by making requests of humans. Developers use the Amazon Mechanical Turk web services API to submit tasks to the Amazon Mechanical Turk web site, approve completed tasks, and incorporate the answers into their software applications. To the application, the transaction looks very much like any remote procedure call—the application sends the request, and the service returns the results. In reality, a network of humans fuels this Artificial Intelligence by coming to the web site, searching for and completing tasks, and receiving payment for their work.
- All software developers need to do is write normal code. The pseudo code below illustrates how simple this can be.
-
read (photo); photoContainsHuman = callMechanicalTurk(photo); if (photoContainsHuman = = TRUE){ acceptPhoto; } else { rejectPhoto; } - More information about Amazon's Mechanical Turk service is provided in the Amazon Mechanical Turk Developer Guide, 2006, 165 pp., API Version 10-31-2006 (attached as Appendix A to priority application Ser. No. 12/114,612).
- The Mechanical Turk service may be regarded as a structured implementation of a technology commonly termed “crowdsourcing”—employing a group of outsiders to perform a task. Wikipedia explains:
-
- “Crowdsourcing” is a neologism for a business model that depends on work being done outside the traditional company walls: while outsourcing is typically performed by lower paid professionals, crowdsourcing relies on a combination of volunteers and low-paid amateurs who use their spare time to create content, solve problems, or even do corporate R&D. The term was coined by Wired magazine writer Jeff Howe and editor Mark Robinson in June 2006.
- Crowds targeted for crowdsourcing include garage scientists, amateur videographers, freelancers, photo enthusiasts, data companies, writers, smart mobs and the electronic herd.
- Overview
-
- While not a new idea, crowdsourcing is becoming mainstream. Open source projects are a form of crowdsourcing that has existed for years. People who may not know one another work together online to create complex software such as the Linux kernel, and the Firefox browser. In recent years internet technology has evolved to allow non-technical people to participate in online projects. Just as important, crowdsourcing presumes that a large number of enthusiasts can outperform a small group of experienced professionals.
- Advantages
-
- The main advantages of crowdsourcing is that innovative ideas can be explored at relatively little cost. Furthermore, it also helps reduce costs. For example if customers reject a particular design, it can easily be scrapped. Though disappointing, this is far less expensive than developing high volumes of a product that no one wants. Crowdsourcing is also related to terms like Collective Customer Commitment (CCC) and Mass Customisation. Collective Customer Commitment (CCC) involves integrating customers into innovation processes. It helps companies exploit a pool of talent and ideas and it also helps firms avoid product flops. Mass Customisation is somewhat similar to collective customer commitment; however, it also helps companies avoid making risky decisions about what components to prefabricate and thus avoids spending for products which may not be marketable later.
- Types of Crowdsourced Work
-
- Steve Jackson Games maintains a network of MIB (Men In Black), who perform secondary jobs (mostly product representation) in exchange for free product. They run publicly or semi-publicly announced play-tests of all their major books and game systems, in exchange for credit and product. They maintain an active user community online, and have done so since the days of BBSes.
- Procter & Gamble employs more than 9000 scientists and researchers in corporate R&D and still have many problems they can't solve. They now post these on a website called InnoCentive, offering large cash rewards to more than 90,000 ‘solvers’ who make up this network of backyard scientists. P&G also works with NineSigma, YourEncore and Yet2.
- Amazon Mechanical Turk co-ordinates the use of human intelligence to perform tasks which computers are unable to do.
- YRUHRN used Amazon Mechanical Turk and other means of crowdsourcing to compile content for a book published just 30 days after the project was started.
- iStockphoto is a website with over 22,000 amateur photographers who upload and distribute stock photographs. Because it does not have the same margins as a professional outfit like Getty Images it is able to sell photos for a low price. It was recently purchased by Getty Images.
- Cambrian House applies a crowdsourcing model to identify and develop profitable software ideas. Using a simple voting model, they attempt to find sticky software ideas that can be developed using a combination of internal and crowdsourced skills and effort.
- A Swarm of Angels is a project to utilize a swarm of subscribers (Angels) to help fund, make, contribute, and distribute, a £1 million feature film using the Internet and all digital technologies. It aims to recruit earlier development community members with the right expertise into paid project members, film crew, and production staff.
- The Goldcorp Challenge is an example of how a traditional company in the mining industry used a crowdsource to identify likely veins of gold on its Red Lake Property. It was won by Fractal Graphics and Taylor-Wall and Associates of Australia but more importantly identified 110 drilling targets, 50% of which were new to the company.
- CafePress and Zazzle, customized products marketplaces for consumers to create apparel, posters, cards, stamps, and other products.
- Marketocracy, to isolating top stock market investors around the world in head to head competition so they can run real mutual funds around these soon-to-be-discovered investment super-stars.
- Threadless, an Internet-based clothing retailer that sells t-shirts which have been designed by and rated by its users.
- Public Insight Journalism, A project at American Public Media to cover the news by tapping the collective and specific intelligence of the public. Gets the newsroom beyond the usual sources, uncovers unexpected expertise, stories and new angles.
- External Links and References
-
- The Rise of Crowdsourcing, Wired June 2006.
- Crowdsourcing: Consumers as Creators, BusinessWeek July 2006.
- In accordance with certain embodiments of this technology, Amazon's Mechanical Turk system, or similar crowdsourcing arrangements, are employed to match a first item of visual or audio entertainment content to a counterpart in a universe of such items.
- For example, consider a user social networking site such as YouTube (now Google) that distributes “user generated content” (e.g., video files), and employs fingerprinting to recognize media content that should not be distributed. The site may check a video file at the time of its uploading with a fingerprint recognition system (e.g., of the sort offered by Audible Magic, or Gracenote). If no clear match is identified, the video may be indexed and stored on YouTube's servers, available for public downloading. Meanwhile, the content can be queued for review by one or more crowdsource reviewers. They may recognize it as a clip from the old TV sitcom “I Love Lucy”—perhaps digitally rotated 3 degrees to avoid fingerprint detection. This tentative identification is returned to YouTube from the API call. YouTube can check the returning metadata against a title list of works that should not be distributed (e.g., per the request of copyright owners), and may discover that “I Love Lucy” clips should not be distributed. It can then remove the content from public distribution. Additionally, the fingerprint database can be updated with the fingerprint of the rotated version of the I Love Lucy clip, allowing it to be immediately recognized the next time it is encountered.
- The foregoing and other examples, features and advantages of this technology will be more apparent from the following Detailed Description.
- One use of the Mechanical Turk service (and similar crowdsourcing technologies) is in connection with computationally difficult tasks, such as identification of audio, video and imagery content. These tasks are sometimes addressed by so-called “fingerprint” technology, which seeks to generate a “robust hash” of content (e.g., distilling a digital file of the content down to perceptually relevant features), and then compare the thus-obtained fingerprint against a database of reference fingerprints computed from known pieces of content, to identify a “best” match. Such technology is detailed, e.g., in Haitsma, et al, “A Highly Robust Audio Fingerprinting System,” Proc. Intl Conf on Music Information Retrieval, 2002; Cano et al, “A Review of Audio Fingerprinting,” Journal of VLSI Signal Processing, 41, 271, 272, 2005; Kalker et al, “Robust Identification of Audio Using Watermarking and Fingerprinting,” in Multimedia Security Handbook, CRC Press, 2005, and in patent documents WO02/065782, US20060075237, US20050259819, and US20050141707.
- A related technology is facial recognition—matching an unknown face to a reference database of facial images. Again, each of the faces is distilled down to a characteristic set of features, and a match is sought between an unknown feature set, and feature sets corresponding to reference images. (The feature set may comprise eigenvectors or shape primitives.) Patent documents particularly concerned with such technology include US20020031253, U.S. Pat. Nos. 6,292,575, 6,301,370, 6,430,306, 6,466,695, and 6,563,950.
- These are examples of technology that relies on “fuzzy” matching. The fingerprint derived from the unknown content often will not exactly match any of the reference fingerprints in the database. Thus, the database must be searched not just for the identical content fingerprint, but also for variants.
- Expanding the search to include variants hugely complicates—and slows—the database search task. To make the search tractable, one approach is to prune the database—identifying excerpts thereof that are believed to be relatively likely to have a match, and limiting the search to those excerpts (or, similarly, identifying excerpts that are believed relatively unlikely to have a match, and not searching those excerpts).
- The database search may locate several reference fingerprints that are similar to the fingerprint of the unknown content. The identification process then seeks to identify a “best” match, using various algorithms.
- Such content identification systems can be improved by injecting a human into the process—by the Mechanical Turk service or similar systems.
- In one particular arrangement, the content identification system makes an assessment of the results of its search, e.g., by a score. A score of 100 may correspond to a perfect match between the unknown fingerprint and a reference fingerprint. Lower scores may correspond to successively less correspondence. (At some lower score, Sx, (perhaps 60) the system may decide that there is no suitable match, and a “no-match” result is returned, with no identification made.)
- Above some threshold score, Sy, (perhaps 70) the system may be sufficiently confident of the result that no human intervention is necessary. At scores below Sy, the system may make a call through the Mechanical Turk service for assistance.
- The Mechanical Turk can be presented the unknown content (or an excerpt thereof), and some reference content, and asked to make a comparison. (The reference content may be stored in the fingerprint database, or may be readily obtainable through use of a link stored in the reference database.)
- A single item of reference content can be provided for comparison with the unknown content, or several items of reference content can be provided. (Again, excerpts may be used instead of the complete content objects. Depending on the application, the content might be processed before sending to the crowdsource engine, e.g., removing metadata (such as personally identifiable information: name, driver license number, etc.) that is printed on, or conveyed with, the file.)
- The requested comparison can take different forms. The service can be asked simply whether two items appear to match. Or it can be asked to identify the best of several possible matches (or indicate that none appears to match). Or it can be asked to give a relative match score (e.g., 0-100) between the unknown content and one or more items reference content.
- In many embodiments, a query is referred to several different humans (e.g., 2-50) through the Mechanical Turk service, and the returned results are examined for consensus on a particular answer. In some queries (e.g., does Content A match Content B? Or is Content A a better match to Content C?), a “vote” may be taken. A threshold of consensus (e.g., 51%, 75%, 90%, 100%) may be required in order for the service response to be given weight in the final analysis. Likewise, in queries that ask the humans to provide a subjective score, the scores returned from plural such calls may be combined to yield a net result. (The high and/or low and/or outlier scores may be disregarded in computing the net result; weighting can sometimes be employed, as noted below.)
- As suggested, the data returned from the Mechanical Turk calls may serve as a biasing factor, e.g., pushing an algorithmically determined output one way or another, to yield a final answer (e.g., a net score). Or the data returned from the Mechanical Turk calls may be treated as a definitive answer—with results from preceding processes disregarded.
- Sometimes the database search may reveal several candidate matches, all with comparable scores (which may be above the threshold Sy). Again, one or more calls to the Mechanical Turk service may be invoked to decide which match is the best, from a subjective human standpoint.
- Sometimes the Mechanical Turk service can be invoked even in situations where the original confidence score is below the threshold, Sx, which is normally taken as indicating “no match.” Thus, the service can be employed to effectively reduce this threshold—continuing to search for potential matches when the rote database search does not yield any results that appear reliable.
- The service can also be invoked to effect database pruning. For example, a database may be organized with several partitions (physical or logical), each containing information of a different class. In a facial recognition database, the data may be segregated by subject gender (i.e., male facial portraits, female facial portraits), and/or by age (15-40, 30-65, 55 and higher—data may sometimes be indexed in two or more classifications), etc. In an image database, the data may be segregated by topical classification (e.g., portrait, sports, news, landscape). In an audio database, the data may be segregated by type (spoken word, music, other). Each classification, in turn, can be further segregated (e.g., “music” may be divided into classical, country, rock, other). And these can be further segregated (e.g., “rock” may be classified by genre, such as soft rock, hard rock, Southern rock; by artist, e.g., Beatles, Rolling Stones, etc.).
- A call to the Mechanical Turk can be made, passing the unknown content object (or an excerpt thereof) to a human reviewer, soliciting advice on classification. The human can indicate the apparent class to which the object belongs (e.g., is this a male or female face? Is this music classical, country, rock, or other?). Or, the human can indicate one or more classes to which the object does not belong.
- With such human advice (which, again, may involve several human reviewers, with a voting or scoring arrangement), the system can focus the database search where a correct match—if any—is more likely to be found (or avoid searching in unproductive database excerpts). This focusing can be done at different times. In one scenario it is done after a rote search is completed, in which the search results yield matches below the desired confidence level of Sy. If the database search space is thereafter restricted by application of human judgment, the search can be conducted again in the limited search space. A more thorough search can be undertaken in the indicated subset(s) of the database. Since a smaller excerpt is being searched, a looser criteria for a “match” might be employed, since the likelihood of false-positive matches is diminished. Thus, for example, the desired confidence level Sy might be reduced from 70 to 65. Or the threshold Sx at which “no match” is concluded, may be reduced from 60 to 55. Alternatively, the focusing can be done before any rote searching is attempted.
- The result of such a human-focused search may reveal one or more candidate matches. The Mechanical Turk service may be called a second time, to vet the candidate matches—in the manner discussed above. This is one of several cases in which it may be desirable to cascade Mechanical Turk calls—the subsequent calls benefiting from the former.
- In the example just-given, the first Mechanical Turk call aids in pruning the database for subsequent search. The second call aids in assessing the results of that subsequent search. In other arrangements, Mechanical Turk calls of the same sort can be cascaded.
- For example, the Mechanical Turk first may be called to identify audio as music/speech/other. A second call may identify music (identified per the first call) as classical/country/rock/other. A third call may identify rock (identified per the second call) as Beatles/Rolling Stones/etc. Here, again, by iterative calling of a crowdsourcing service, a subjective judgment can be made that would be very difficult to achieve otherwise.
- In some arrangements, human reviewers are pre-qualified as knowledgeable in a specific domain (e.g., relatively expert in recognizing Beatles music). This qualification can be established by an online examination, which reviewers are invited to take to enable them to take on specific tasks (often at an increased rate of pay). Some queries may be routed only to individuals that are pre-qualified in a particular knowledge domain. In the cascaded example just given, for example, the third call might be routed to one or more users with demonstrated expertise with the Beatles (and, optionally, to one or more users with demonstrated expertise with the Rolling Stones, etc.). A positive identification of the unknown content as sounding like the Beatles would be given more relative weight if coming from a human qualified in this knowledge domain. (Such weighting may be taken into account when aggregating results from plural human reviewers. For example, consider an unknown audio clip sent to six reviewers, two with expertise in the Beatles, two with expertise in the Rolling Stones, and two with expertise in the Grateful Dead. Assume the Beatles experts identify it as Beatles music, the Rolling Stones experts identify it as Grateful Dead music, and the Grateful Dead experts identify it as Rolling Stones music. Despite the fact that there are tie votes, and despite the fact that no selection earned a majority of the votes, the content identification service that made these calls and is provided with these results may logically conclude that the music is Beatles.)
- Calls to the Mechanical Turk service may request the human to provide metadata relevant to any content reviewed. This can include supposed artist(s), genre, title, subject, date, etc. This information (which may be ancillary to a main request, or may comprise the entirety of the request) can be entered into a database. For example, it can be entered into a fingerprint database—in association with the content reviewed by the human.
- Desirably, data gleaned from Mechanical Turk calls are entered into the database, and employed to enrich its data—and enrich information that can be later mined from the database. For example, if unknown content X has a fingerprint Fx, and through the Mechanical Turk service it is determined that this content is a match to reference content Y, with fingerprint Fy, then a corresponding notation can be added to the database, so that a later query on fingerprint Fx (or close variants thereof) will indicate a match to content Y. (E.g., a lookup table initially indexed with a hash of the fingerprint Fx will point to the database record for content Y.)
- Calls to outsourcing engines involve a time lag before results are returned. The calling system can generally cope, or be adapted to cope, with such lags.
- Consider a user social networking site such as YouTube (now Google) that distributes “user generated content” (e.g., video files), and employs fingerprinting to recognize media content that should not be distributed. The site may check a video file at the time of its uploading with a fingerprint recognition system (e.g., of the sort offered by Audible Magic, or Gracenote). If no clear match is identified, the video may be indexed and stored on YouTube's servers, available for public downloading. Meanwhile, the content can be queued for review by one or more crowdsource reviewers. They may recognize it as a clip from the old TV sitcom “I Love Lucy”—perhaps digitally rotated 3 degrees to avoid fingerprint detection. This tentative identification is returned to YouTube from the API call. YouTube can check the returning metadata against a title list of works that should not be distributed (e.g., per the request of copyright owners), and may discover that “I Love Lucy” clips should not be distributed. It can then remove the content from public distribution. (This generally follows a double-check of the identification by a YouTube employee.) Additionally, the fingerprint database can be updated with the fingerprint of the rotated version of the I Love Lucy clip, allowing it to be immediately recognized the next time it is encountered.
- If the content is already being delivered to a user at the moment the determination is made (i.e., the determination that the content should not be distributed publicly), then the delivery can be interrupted. An explanatory message can be provided to the user (e.g., a splash screen presented at the interruption point in the video).
- Rotating a video by a few degrees is one of several hacks that can defeat fingerprint identification. (It is axiomatic that introduction of any new content protection technology draws hacker scrutiny. Familiar examples include attacks against Macrovision protection for VHS tapes, and against CSS protection for packaged DVD discs.) If fingerprinting is employed in content protection applications, such as in social networking sites (as outlined above) or peer-to-peer networks, its vulnerability to attack will eventually be determined and exploited.
- Each fingerprinting algorithm has particular weaknesses that can be exploited by hackers to defeat same. An example will help illustrate.
- A well-known fingerprinting algorithm operates by repeatedly analyzing the frequency content of a short excerpt of an audio track (e.g., 0.4 seconds). The method determines the relative energy of this excerpt within 33 narrow frequency bands that logarithmically span the range 300 Hz-2000 Hz. A corresponding 32-bit identifier is then generated from the resulting data. In particular, a frequency band corresponds to a data bit “1” if its energy level is larger than that of the band above, and a “0” if its energy level is lower. (A more complex arrangement can also take variations over time into account, outputting a “1” only if the immediately preceding excerpt also met the same test, i.e., having a band energy greater than the band above.)
- Such a 32 bit identifier is computed every hundredth of a second or so, for the immediately preceding 0.4 second excerpt of the audio track, resulting in a large number of “fingerprints.” This series of characteristic fingerprints can be stored in a database entry associated with the track, or only a subset may be stored (e.g., every fourth fingerprint).
- When an unknown track is encountered, the same calculation process is repeated. The resulting set of data is then compared against data earlier stored in the database to try and identify a match. (As noted, various strategies can be employed to speed the search over a brute-force search technique, which yields unacceptable search times.) While the just-described technique is designed for audio identification, a similar arrangement can be used for video. Instead of energies in audio subbands, the algorithm can use average luminances of blocks into which the image is divided as the key perceptual features. Again, a fingerprint can be defined by determining whether the luminance in each block is larger or smaller than the luminance of the preceding block.
- While little has been written about attacks targeting fingerprinting systems, a casual examination of possible attack scenarios reveals several possibilities. A true hacker will probably see many more. Four simple approaches are discussed below.
- The reader may be familiar with different loudness profiles selectable on car radios, e.g., Jazz, Talk, Rock, etc. Each applies a different frequency equalization profile to the audio, e.g., making bass notes louder if the Rock setting is selected, and quieter if the Talk setting is selected, etc. The difference is often quite audible when switching between different settings.
- However, if the radio is simply turned on and tuned to different stations, the listener is generally unaware of which loudness profile is being employed. That is, without the ability to switch between different profiles, the frequency equalization imposed by a particular loudness profile is typically not noticed by a listener. The different loudness profiles, however, yield different fingerprints.
- For example, in the Rock setting, the 300 Hz energy in a particular 0.4 second excerpt may be greater than the 318 Hz energy. However, in the Talk setting, the situation may be reversed. This change prompts a change in the leading bit of the fingerprint.
- In practice, an attacker would probably apply loudness profiles more complex than those commonly available in car radios—increasing and decreasing the loudness at many different frequency bands (e.g., 32 different frequency bands). Significantly different fingerprints may thus be produced. Moreover, the loudness profile could change with time—further distancing the resulting fingerprint from the reference values stored in a database.
- Another process readily available to attackers is audio multiband compression, a form of processing that is commonly employed by broadcasters to increase the apparent loudness of their signal (most especially commercials). Such tools operate by reducing the dynamic range of a soundtrack—increasing the loudness of quiet passages on a band-by-band basis, to thereby achieve a higher average signal level. Again, this processing of the audio changes its fingerprint, yet is generally not objectionable to the listeners.
- The two examples given above are informal attacks—common signal processing techniques that yield, as side-effects, changes in audio fingerprints. Formal attacks—signal processing techniques that are optimized for purposes of changing fingerprints—are numerous.
- Some formal attacks are based on psychoacoustic masking. This is the phenomena by which, e.g., a loud sound at one instant (e.g., a drum beat) obscures a listener's ability to perceive a quieter sound at a later instant. Or the phenomena by which a loud sound at one frequency (e.g., 338 Hz) obscures a listener's ability to perceive a quieter sound at a nearby frequency (e.g., 358 Hz) at the same instant. Research in this field goes back decades. (Modern watermarking software employs psychoacoustic masking in an advantageous way, to help hide extra data in audio and video content.)
- Hacking software, of course, can likewise examine a song's characteristics and identify the psychoacoustic masking opportunities it presents. Such software can then automatically make slight alterations in the song's frequency components in a way that a listener won't be able to note, yet in a way that will produce a different series of characteristic fingerprints. The processed song will be audibly indistinguishable from the original, but will not “match” any series of fingerprints in the database.
- Another formal attack targets fingerprint bit determinations that are near a threshold, and slightly adjusts the signal to swing the outcome the other way. Consider an audio excerpt that has the following respective energy levels (on a scale of 0-99), in the frequency bands indicated:
-
300 Hz 318 Hz 338 Hz 358 Hz 69 71 70 68 - The algorithm detailed above would generate a fingerprint of {011 . . . } from this data (i.e., 69 is less than 71, so the first bit is ‘0’; 71 is greater than 70, so the second bit is ‘1’; 70 is greater than 68, so the third bit is ‘1’).
- Seeing that the energy levels are somewhat close, an attacker tool could slightly adjust the signal's spectral composition, so that the relative energy levels are as follows:
-
300 Hz 318 Hz 338 Hz 358 Hz [69] 70 [71] 69 70 68 - Instead of {011 . . . }, the fingerprint is now {101 . . . }. Two of the three illustrated fingerprint bits have been changed. Yet the change to the audio excerpt is essentially inaudible.
- Other fingerprint hacking vulnerabilities arise from shortcuts employed in the database searching strategy—seeking to prune large segments of the data from further searching. For example, the system outlined above confines the large potential search space by assuming that there exists a 32 bit excerpt of the unknown song fingerprint that exactly matches (or matches with only one bit error) a 32 bit excerpt of fingerprint data in the reference database. The system looks at successive 32 bit excerpts from the unknown song fingerprint, and identifies all database fingerprints that include an excerpt presenting a very close match (i.e., 0 or 1 errors). A list of candidate song fingerprints is thereby identified that can be further checked to determine if any meets the looser match criteria generally used. (To allow non-exact fingerprint matches, the system generally allows up to 2047 bit errors in every 8192 bit block of fingerprint data.)
- The evident problem is: what if the correct “match” in the database has no 32 bit excerpt that corresponds—with just 1 or 0 bit errors—to a 32 bit excerpt from the unknown song? Such a correct match will never be found—it gets screened out at the outset.
- A hacker familiar with the system's principles will see that everything hinges on the assumption that a 32 bit string of fingerprint data will identically match (or match with only one bit error) a corresponding string in the reference database. Since these 32 bits are based on the strengths of 32 narrow frequency bands between 300 Hz and 2000 Hz, the spectrum of the content can readily be tweaked to violate this assumption, forcing a false-negative error. (E.g., notching out two of these narrow bands will force four bits of every 32 to a known state: two will go to zero—since these bands are lower in amplitude than the preceding bands, and two will go to one—since the following bands are higher in amplitude that these preceding, notched, bands). On average, half of these forced bits will be “wrong” (compared to the untweaked music), leading to two bit errors—violating the assumption on which database pruning is based.)
- Attacks like the foregoing require a bit of effort. However, once an attacker makes the effort, the resulting hack can be spread quickly and widely.
- The exemplary fingerprinting technique noted above (which is understood to be the basis for Gracenote's commercial implementation, MusicID, built from technology licensed from Philips) is not unique in being vulnerable to various attacks. All fingerprinting techniques (including the recently announced MediaHedge, as well as CopySense and RepliCheck) are similarly believed to have vulnerabilities that can be exploited by hackers. (A quandary for potential adopters is that susceptibility of different techniques to different attacks has not been a focus of academic attention.)
- It will be recognized that crowdsourcing can help mitigate the vulnerabilities and uncertainties that are inherent in fingerprinting systems. Despite a “no-match” returned from the fingerprint-based content identification system (based on its rote search of the database for a fingerprint that matches that of the altered content), the techniques detailed herein allow human judgment to take a “second look.” Such techniques can identify content that has been altered to avoid its correct identification by fingerprint techniques. (Again, once such identification is made, corresponding information is desirably entered into the database to facilitate identification of the altered content next time.)
- It will be recognized that the “crowdsourcing” methodologies detailed above also have applicability to other tasks involved in the arrangements detailed in this specification, including all the documents incorporated by reference.
- Implementation of systems according to the present technology is straightforward to artisans, e.g., using standard computer-, database-, software- and network-technology.
- To provide a comprehensive disclosure without unduly lengthening this specification, applicant incorporates-by-reference the documents referenced in this disclosure. In addition to those noted elsewhere, these incorporated documents include application Ser. No. 10/979,770 (now U.S. Pat. No. 7,314,162) and Ser. No. 11/132,724 (published as US20050288952); published applications US20030052768, US20030099379. US20030115459, US20030216988, US20040059953, US20040064415, US20040153663, US20040189441, US20040205030, US20040213437, US20040230527, US20040245330, US20050039057, US20050132235, US20050154924, and US20050171851, and issued U.S. Pat. Nos. 5,679,938, 5,679,940, 6,513,018, 6,597,775, 6,944,604, 6,965,889, and 6,968,328.
- It is expressly contemplated that the technologies, features and analytical methods detailed herein can be incorporated into the methods/systems detailed in such other documents. Moreover, the technologies, features, and analytical methods detailed in those documents can be incorporated into the methods/systems detailed herein. (It will be recognized that the brief synopses of prior documents provided above naturally do not reflect all of the features found in such disclosures.)
- In view of the wide variety of embodiments to which the principles and features discussed above can be applied, it should be apparent that the detailed embodiments are illustrative only and should not be taken as limiting the scope of the disclosed technology. Rather, we claim all such modifications as may come within the scope and spirit of the following claims and equivalents thereof.
Claims (23)
1-8. (canceled)
9. A computer-implemented method involving visual or audio entertainment content uploaded to a first party's video service for distribution from said service to content viewers, the uploaded content including a portion apparently derived from a reference work, the method being executed by one or more computer systems and comprising:
selecting a set of stored rule data, from among plural sets of stored rule data, that should govern treatment of the uploaded content, said selecting being based on the reference work from which a portion of the uploaded content apparently was derived, or based on a proprietor of said reference work;
determining an action to take concerning the uploaded content, based on the selected set of stored rule data, and also based on one or more input factor data, wherein said input factor data comprises factor data selected from a list consisting of: (a) a time length of said derived portion, (b) a fraction of the reference work that corresponds to said derived portion, and (c) a part of the reference work from which said portion apparently was derived; and
taking said determined action.
10. The method of claim 9 that includes determining the action to take concerning the uploaded content based on at least factor (b) or (c) from said list.
11. The method of claim 9 that includes determining the action to take concerning the uploaded content based on at least two factors from said list.
12. The method of claim 9 that includes performing the method repeatedly, with different uploaded content, and different reference works from which portions of the uploaded content apparently were derived, wherein:
in a first circumstance, the determined action comprises notifying a reference work proprietor about the uploaded content;
in a second circumstance, the determined action comprises preventing the uploaded content from being distributed from the video service to content viewers; and
in a third circumstance, the determined action comprises sharing, with a reference work proprietor, revenue generated from advertising presented to content viewers in connection with the uploaded content.
13. A computer-implemented method involving visual or audio entertainment content uploaded to a first party's video service for distribution from said service to content viewers, the method being executed by one or more computer systems and comprising:
identifying a portion of the uploaded content that apparently was derived from a reference work;
determining one or more factors selected from a list consisting of: (a) a time length of said derived portion, (b) a fraction of the reference work that corresponds to said derived portion, and (c) a part of the reference work from which said portion apparently was derived; and
taking an action concerning the uploaded content based at least in part on said determined factor(s).
14. The method of claim 13 in which the determining comprises determining at least factor (a) from said list.
15. The method of claim 14 in which said determining additionally comprises determining factor (c) from said list.
16. The method of claim 13 in which the determining comprises determining at least factor (b) from said list.
17. The method of claim 16 in which said determining additionally comprises determining factor (c) from said list.
18. The method of claim 13 in which the determining comprises determining at least factor (c) from said list.
19. The method of claim 13 wherein said determining proceeds with reference to received fingerprint data computed from the uploaded content, and reference fingerprint data that corresponds to the reference work.
20. The method of claim 13 that further includes determining a confidence score indicating confidence that said portion of the uploaded content was derived from said reference work, wherein said action additionally depends on said confidence score.
21. The method of claim 20 in which:
said action includes identifying the uploaded content and the reference work to a second party;
and in which the method further includes:
collecting human assessment data from said second party, the assessment data indicating a result of a comparison of at least a portion of the uploaded content with at least a portion of the reference work by one or more human reviewer(s); and
taking a further action based at least in part on said collected human assessment data.
22. The method of claim 13 that includes consulting stored rules to determine what action to take, based at least in part on said determined factor(s).
23. The method of claim 22 that includes consulting first stored rule data when the reference work is associated with a first proprietor, and consulting second, different, stored rule data when the reference work is associated with a second proprietor.
24. The method of claim 23 wherein the first stored rule data comprises default rules used because the first proprietor did not select alternate rules, and wherein the second stored rule data comprises alternate rules selected by the second proprietor.
25. The method of claim 22 , performed repeatedly with different uploaded content, and different reference works, wherein the method includes consulting a first set of stored rules when the reference work comprises a first reference work, and consulting a second, different set of stored rules when the reference work comprises a second reference work different than the first reference work.
26. The method of claim 25 wherein the first set of stored rules comprises a default set of rules defined by the first party.
27. The method of claim 22 , performed repeatedly with different uploaded content, and different reference works, wherein:
in a first circumstance, the action comprises notifying a proprietor of the reference work about the uploaded content;
in a second circumstance, the action comprises preventing the uploaded content from being distributed from the video service to content viewers; and
in a third circumstance, the action comprises sharing, with a proprietor of the reference work, revenue generated from advertising presented to content viewers in connection with the uploaded content.
28. The method of claim 13 in which said action also depends on a geographic location of a user that uploaded the content to the video service or that requested download of the content from the video service.
29. A computer-implemented method involving visual or audio entertainment content uploaded to a first party's video service for distribution from said service to content viewers, the method being executed by one or more computer systems and comprising:
analyzing fingerprint data corresponding to the uploaded content, and fingerprint data corresponding to reference content, to identify a portion of the uploaded content that apparently was derived from a work of reference content;
selecting stored rule data that should govern treatment of the uploaded content, from among plural sets of stored rule data, said selecting being based on said reference work or based on a proprietor of said reference work;
determining plural factors including (a) a length of said portion of the uploaded content that apparently was derived from the reference work, (b) from what part of the reference work was said portion of the uploaded content apparently derived, and (c) a confidence score indicating confidence that said portion of the uploaded content was derived from said reference work;
determining an action to take concerning the uploaded content, based on two or more of said plural factors, and also based on the selected stored rule data;
wherein the determined action comprises:
identifying the uploaded content and the reference work to a second party;
collecting human assessment data from said second party, the assessment data indicating a result of a comparison of at least a portion of the uploaded content with at least a portion of the reference work by one or more human reviewer(s); and
determining a further action to take concerning the uploaded content based at least in part on said collected human assessment data.
30. The method of claim 29 , performed repeatedly with different uploaded content, and different reference works, wherein the method includes determining the action to take based on a first set of stored rules when the reference work comprises a first reference work having a first proprietor, and determining the action to take based on a second, different set of stored rules when the reference work comprises a second, different, reference work having a second, different, proprietor, wherein the first set of stored rules comprises a default set of rules defined by the first party, and the second set of store rules comprises an alternate set of rules selected by the second proprietor.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US16/360,896 US20200058088A1 (en) | 2007-03-14 | 2019-03-21 | Method and system for determining content treatment |
Applications Claiming Priority (5)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US89485907P | 2007-03-14 | 2007-03-14 | |
US12/048,072 US9179200B2 (en) | 2007-03-14 | 2008-03-13 | Method and system for determining content treatment |
US12/114,612 US8341412B2 (en) | 2005-12-23 | 2008-05-02 | Methods for identifying audio or video content |
US13/686,541 US10242415B2 (en) | 2006-12-20 | 2012-11-27 | Method and system for determining content treatment |
US16/360,896 US20200058088A1 (en) | 2007-03-14 | 2019-03-21 | Method and system for determining content treatment |
Related Parent Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US13/686,541 Continuation US10242415B2 (en) | 2006-12-20 | 2012-11-27 | Method and system for determining content treatment |
Publications (1)
Publication Number | Publication Date |
---|---|
US20200058088A1 true US20200058088A1 (en) | 2020-02-20 |
Family
ID=69523275
Family Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US13/686,541 Expired - Fee Related US10242415B2 (en) | 2006-12-20 | 2012-11-27 | Method and system for determining content treatment |
US16/360,896 Abandoned US20200058088A1 (en) | 2007-03-14 | 2019-03-21 | Method and system for determining content treatment |
Family Applications Before (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US13/686,541 Expired - Fee Related US10242415B2 (en) | 2006-12-20 | 2012-11-27 | Method and system for determining content treatment |
Country Status (1)
Country | Link |
---|---|
US (2) | US10242415B2 (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20220027407A1 (en) * | 2020-07-27 | 2022-01-27 | Audible Magic Corporation | Dynamic identification of unknown media |
Families Citing this family (21)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20070162761A1 (en) | 2005-12-23 | 2007-07-12 | Davis Bruce L | Methods and Systems to Help Detect Identity Fraud |
US8738749B2 (en) | 2006-08-29 | 2014-05-27 | Digimarc Corporation | Content monitoring and host compliance evaluation |
US8707459B2 (en) | 2007-01-19 | 2014-04-22 | Digimarc Corporation | Determination of originality of content |
US8010511B2 (en) | 2006-08-29 | 2011-08-30 | Attributor Corporation | Content monitoring and compliance enforcement |
US10977693B2 (en) * | 2008-11-26 | 2021-04-13 | Free Stream Media Corp. | Association of content identifier of audio-visual data with additional data through capture infrastructure |
US20150019653A1 (en) * | 2013-07-15 | 2015-01-15 | Civolution B.V. | Method and system for adding an identifier |
JP2015037212A (en) * | 2013-08-12 | 2015-02-23 | オリンパスイメージング株式会社 | Information processing device, imaging equipment and information processing method |
US10824747B1 (en) | 2017-01-25 | 2020-11-03 | State Farm Mutual Automobile Insurance Company | Systems and methods for controlled access to policy data on blockchain |
CN106960290B (en) * | 2017-04-11 | 2023-12-22 | 西华大学 | System and method for evaluating sales service quality of automobile 4S shop team |
US20180300699A1 (en) * | 2017-04-12 | 2018-10-18 | Facebook, Inc. | Systems and methods for content management |
US9936230B1 (en) * | 2017-05-10 | 2018-04-03 | Google Llc | Methods, systems, and media for transforming fingerprints to detect unauthorized media content items |
US10015546B1 (en) * | 2017-07-27 | 2018-07-03 | Global Tel*Link Corp. | System and method for audio visual content creation and publishing within a controlled environment |
CN107731220B (en) * | 2017-10-18 | 2019-01-22 | 北京达佳互联信息技术有限公司 | Audio identification methods, device and server |
WO2019087282A1 (en) * | 2017-10-31 | 2019-05-09 | Line株式会社 | Information processing method, information processing device and program |
US20190199519A1 (en) * | 2017-12-26 | 2019-06-27 | Microsoft Technology Licensing, Llc | Detecting and treating unauthorized duplicate digital content |
WO2020123124A1 (en) * | 2018-12-14 | 2020-06-18 | Google Llc | Methods, systems, and media for identifying videos containing objectionable content |
US20220237637A1 (en) * | 2018-12-18 | 2022-07-28 | Meta Platforms, Inc. | Systems and methods for real time crowdsourcing |
US11106827B2 (en) | 2019-03-26 | 2021-08-31 | Rovi Guides, Inc. | System and method for identifying altered content |
US11134318B2 (en) * | 2019-03-26 | 2021-09-28 | Rovi Guides, Inc. | System and method for identifying altered content |
CN111741231B (en) * | 2020-07-23 | 2022-02-22 | 北京字节跳动网络技术有限公司 | Video dubbing method, device, equipment and storage medium |
CN112381355A (en) * | 2020-10-23 | 2021-02-19 | 广东电网有限责任公司中山供电局 | Unattended data processing system and method for safety tool warehouse |
Family Cites Families (153)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5210820A (en) | 1990-05-02 | 1993-05-11 | Broadcast Data Systems Limited Partnership | Signal recognition system and method |
US8505108B2 (en) * | 1993-11-18 | 2013-08-06 | Digimarc Corporation | Authentication using a digital watermark |
US6122403A (en) | 1995-07-27 | 2000-09-19 | Digimarc Corporation | Computer system linked by using information in data objects |
US6947571B1 (en) | 1999-05-19 | 2005-09-20 | Digimarc Corporation | Cell phones with optical capabilities, and related applications |
US5715403A (en) | 1994-11-23 | 1998-02-03 | Xerox Corporation | System for controlling the distribution and use of digital works having attached usage rights where the usage rights are defined by a usage rights grammar |
JPH08263438A (en) * | 1994-11-23 | 1996-10-11 | Xerox Corp | Distribution and use control system of digital work and access control method to digital work |
US7133846B1 (en) | 1995-02-13 | 2006-11-07 | Intertrust Technologies Corp. | Digital certificate support system, methods and techniques for secure electronic commerce transaction and rights management |
US7124302B2 (en) * | 1995-02-13 | 2006-10-17 | Intertrust Technologies Corp. | Systems and methods for secure transaction management and electronic rights protection |
US7711564B2 (en) | 1995-07-27 | 2010-05-04 | Digimarc Corporation | Connected audio and other media objects |
US5765152A (en) | 1995-10-13 | 1998-06-09 | Trustees Of Dartmouth College | System and method for managing copyrighted electronic media |
US7047241B1 (en) | 1995-10-13 | 2006-05-16 | Digimarc Corporation | System and methods for managing digital creative works |
US7159116B2 (en) | 1999-12-07 | 2007-01-02 | Blue Spike, Inc. | Systems, methods and devices for trusted transactions |
US6108637A (en) | 1996-09-03 | 2000-08-22 | Nielsen Media Research, Inc. | Content display monitor |
US6647548B1 (en) | 1996-09-06 | 2003-11-11 | Nielsen Media Research, Inc. | Coded/non-coded program audience measurement system |
US20020120925A1 (en) | 2000-03-28 | 2002-08-29 | Logan James D. | Audio and video program recording, editing and playback systems using metadata |
US5983351A (en) | 1996-10-16 | 1999-11-09 | Intellectual Protocols, L.L.C. | Web site copyright registration system and method |
US6251016B1 (en) | 1997-01-07 | 2001-06-26 | Fujitsu Limited | Information offering system for providing a lottery on a network |
AU729891B2 (en) | 1997-02-07 | 2001-02-15 | About.Com Inc. | Collaborative internet data mining system |
US6055538A (en) | 1997-12-22 | 2000-04-25 | Hewlett Packard Company | Methods and system for using web browser to search large collections of documents |
US7051004B2 (en) | 1998-04-03 | 2006-05-23 | Macrovision Corporation | System and methods providing secure delivery of licenses and content |
US7756892B2 (en) | 2000-05-02 | 2010-07-13 | Digimarc Corporation | Using embedded data with file sharing |
US7689532B1 (en) | 2000-07-20 | 2010-03-30 | Digimarc Corporation | Using embedded data with file sharing |
US6401118B1 (en) | 1998-06-30 | 2002-06-04 | Online Monitoring Services | Method and computer program product for an online monitoring search engine |
US6603921B1 (en) | 1998-07-01 | 2003-08-05 | International Business Machines Corporation | Audio/video archive system and method for automatic indexing and searching |
US7185201B2 (en) | 1999-05-19 | 2007-02-27 | Digimarc Corporation | Content identifiers triggering corresponding responses |
US7302574B2 (en) | 1999-05-19 | 2007-11-27 | Digimarc Corporation | Content identifiers triggering corresponding responses through collaborative processing |
US7762453B2 (en) | 1999-05-25 | 2010-07-27 | Silverbrook Research Pty Ltd | Method of providing information via a printed substrate with every interaction |
US7284255B1 (en) | 1999-06-18 | 2007-10-16 | Steven G. Apel | Audience survey system, and system and methods for compressing and correlating audio signals |
US7346605B1 (en) | 1999-07-22 | 2008-03-18 | Markmonitor, Inc. | Method and system for searching and monitoring internet trademark usage |
US6493744B1 (en) | 1999-08-16 | 2002-12-10 | International Business Machines Corporation | Automatic rating and filtering of data files for objectionable content |
US6546135B1 (en) | 1999-08-30 | 2003-04-08 | Mitsubishi Electric Research Laboratories, Inc | Method for representing and comparing multimedia content |
US7194752B1 (en) | 1999-10-19 | 2007-03-20 | Iceberg Industries, Llc | Method and apparatus for automatically recognizing input audio and/or video streams |
US7174293B2 (en) | 1999-09-21 | 2007-02-06 | Iceberg Industries Llc | Audio identification system and method |
AU7564200A (en) | 1999-09-22 | 2001-04-24 | Oleg Kharisovich Zommers | Interactive personal information system and method |
US6807634B1 (en) | 1999-11-30 | 2004-10-19 | International Business Machines Corporation | Watermarks for customer identification |
US20020002586A1 (en) | 2000-02-08 | 2002-01-03 | Howard Rafal | Methods and apparatus for creating and hosting customized virtual parties via the internet |
US7298864B2 (en) | 2000-02-19 | 2007-11-20 | Digimarc Corporation | Digital watermarks as a gateway and control mechanism |
US20060080200A1 (en) | 2000-04-07 | 2006-04-13 | Ashton David M | System and method for benefit plan administration |
US6952769B1 (en) | 2000-04-17 | 2005-10-04 | International Business Machines Corporation | Protocols for anonymous electronic communication and double-blind transactions |
US6684254B1 (en) | 2000-05-31 | 2004-01-27 | International Business Machines Corporation | Hyperlink filter for “pirated” and “disputed” copyright material on the internet in a method, system and program |
US6990453B2 (en) | 2000-07-31 | 2006-01-24 | Landmark Digital Services Llc | System and methods for recognizing sound and music signals in high noise and distortion |
AU2001290581A1 (en) | 2000-08-28 | 2002-03-13 | Digitalowl.Com, Inc. | System and methods for the production, distribution and flexible usage of electronic content in heterogeneous distributed environments |
WO2002019717A2 (en) | 2000-08-31 | 2002-03-07 | Myrio Corporation | Real-time audience monitoring, content rating, and content enhancing |
US20040010417A1 (en) | 2000-10-16 | 2004-01-15 | Ariel Peled | Method and apparatus for supporting electronic content distribution |
US7043473B1 (en) | 2000-11-22 | 2006-05-09 | Widevine Technologies, Inc. | Media tracking system and method |
US7266704B2 (en) | 2000-12-18 | 2007-09-04 | Digimarc Corporation | User-friendly rights management systems and methods |
ATE405101T1 (en) | 2001-02-12 | 2008-08-15 | Gracenote Inc | METHOD FOR GENERATING AN IDENTIFICATION HASH FROM THE CONTENTS OF A MULTIMEDIA FILE |
US20020168082A1 (en) * | 2001-03-07 | 2002-11-14 | Ravi Razdan | Real-time, distributed, transactional, hybrid watermarking method to provide trace-ability and copyright protection of digital content in peer-to-peer networks |
US7681032B2 (en) | 2001-03-12 | 2010-03-16 | Portauthority Technologies Inc. | System and method for monitoring unauthorized transport of digital content |
US7987510B2 (en) | 2001-03-28 | 2011-07-26 | Rovi Solutions Corporation | Self-protecting digital content |
WO2002082271A1 (en) | 2001-04-05 | 2002-10-17 | Audible Magic Corporation | Copyright detection and protection system and method |
US6996273B2 (en) | 2001-04-24 | 2006-02-07 | Microsoft Corporation | Robust recognizer of perceptually similar content |
US20020174132A1 (en) | 2001-05-04 | 2002-11-21 | Allresearch, Inc. | Method and system for detecting unauthorized trademark use on the internet |
US20030061490A1 (en) | 2001-09-26 | 2003-03-27 | Abajian Aram Christian | Method for identifying copyright infringement violations by fingerprint detection |
US7117513B2 (en) | 2001-11-09 | 2006-10-03 | Nielsen Media Research, Inc. | Apparatus and method for detecting and correcting a corrupted broadcast time code |
US7840488B2 (en) | 2001-11-20 | 2010-11-23 | Contentguard Holdings, Inc. | System and method for granting access to an item or permission to use an item based on configurable conditions |
US7477739B2 (en) | 2002-02-05 | 2009-01-13 | Gracenote, Inc. | Efficient storage of fingerprints |
DE10216261A1 (en) * | 2002-04-12 | 2003-11-06 | Fraunhofer Ges Forschung | Method and device for embedding watermark information and method and device for extracting embedded watermark information |
US6885757B2 (en) * | 2002-04-18 | 2005-04-26 | Sarnoff Corporation | Method and apparatus for providing an asymmetric watermark carrier |
US7120273B2 (en) | 2002-05-31 | 2006-10-10 | Hewlett-Packard Development Company, Lp. | Apparatus and method for image group integrity protection |
US8601504B2 (en) | 2002-06-20 | 2013-12-03 | Verance Corporation | Secure tracking system and method for video program content |
US20050259819A1 (en) | 2002-06-24 | 2005-11-24 | Koninklijke Philips Electronics | Method for generating hashes from a compressed multimedia content |
US6931413B2 (en) | 2002-06-25 | 2005-08-16 | Microsoft Corporation | System and method providing automated margin tree analysis and processing of sampled data |
US20040091111A1 (en) | 2002-07-16 | 2004-05-13 | Levy Kenneth L. | Digital watermarking and fingerprinting applications |
SE0202451D0 (en) | 2002-08-15 | 2002-08-15 | Ericsson Telefon Ab L M | Flexible Sim-Based DRM agent and architecture |
US20040044571A1 (en) | 2002-08-27 | 2004-03-04 | Bronnimann Eric Robert | Method and system for providing advertising listing variance in distribution feeds over the internet to maximize revenue to the advertising distributor |
US6983280B2 (en) | 2002-09-13 | 2006-01-03 | Overture Services Inc. | Automated processing of appropriateness determination of content for search listings in wide area network searches |
US8041719B2 (en) | 2003-05-06 | 2011-10-18 | Symantec Corporation | Personal computing device-based mechanism to detect preselected data |
MXPA05004231A (en) | 2002-10-23 | 2005-07-05 | Nielsen Media Res Inc | Digital data insertion apparatus and methods for use with compressed audio/video data. |
JP4366916B2 (en) | 2002-10-29 | 2009-11-18 | 富士ゼロックス株式会社 | Document confirmation system, document confirmation method, and document confirmation program |
KR20050086470A (en) | 2002-11-12 | 2005-08-30 | 코닌클리케 필립스 일렉트로닉스 엔.브이. | Fingerprinting multimedia contents |
JP2004180278A (en) | 2002-11-15 | 2004-06-24 | Canon Inc | Information processing apparatus, server device, electronic data management system, information processing system, information processing method, computer program, and computer-readable storage medium |
US8332326B2 (en) | 2003-02-01 | 2012-12-11 | Audible Magic Corporation | Method and apparatus to identify a work received by a processing system |
US7606790B2 (en) | 2003-03-03 | 2009-10-20 | Digimarc Corporation | Integrating and enhancing searching of media content and biometric databases |
DE10319771B4 (en) * | 2003-05-02 | 2005-03-17 | Koenig & Bauer Ag | System for inspecting a printed image |
JP2005004728A (en) | 2003-05-20 | 2005-01-06 | Canon Inc | Information processing system, information processing device, information processing method, storage medium storing program for executing same so that program can be read out to information processing device, and program |
US7454061B2 (en) | 2003-06-27 | 2008-11-18 | Ricoh Company, Ltd. | System, apparatus, and method for providing illegal use research service for image data, and system, apparatus, and method for providing proper use research service for image data |
JP2007535718A (en) | 2003-07-07 | 2007-12-06 | クリプターグラフィー リサーチ インコーポレイテッド | Reprogrammable security to regulate piracy and enable interactive content |
US20050012563A1 (en) | 2003-07-16 | 2005-01-20 | Michael Kramer | Method and system for the simulataneous recording and identification of audio-visual material |
US20050043548A1 (en) | 2003-08-22 | 2005-02-24 | Joseph Cates | Automated monitoring and control system for networked communications |
US20050049868A1 (en) * | 2003-08-25 | 2005-03-03 | Bellsouth Intellectual Property Corporation | Speech recognition error identification method and system |
US7369677B2 (en) | 2005-04-26 | 2008-05-06 | Verance Corporation | System reactions to the detection of embedded watermarks in a digital host content |
US9055239B2 (en) | 2003-10-08 | 2015-06-09 | Verance Corporation | Signal continuity assessment using embedded watermarks |
US7444403B1 (en) | 2003-11-25 | 2008-10-28 | Microsoft Corporation | Detecting sexually predatory content in an electronic communication |
US8700533B2 (en) | 2003-12-04 | 2014-04-15 | Black Duck Software, Inc. | Authenticating licenses for legally-protectable content based on license profiles and content identifiers |
US7707039B2 (en) | 2004-02-15 | 2010-04-27 | Exbiblio B.V. | Automatic modification of web pages |
US20050193016A1 (en) | 2004-02-17 | 2005-09-01 | Nicholas Seet | Generation of a media content database by correlating repeating media content in media streams |
US7751805B2 (en) | 2004-02-20 | 2010-07-06 | Google Inc. | Mobile image-based information retrieval system |
JP4665406B2 (en) | 2004-02-23 | 2011-04-06 | 日本電気株式会社 | Access control management method, access control management system, and terminal device with access control management function |
US8255331B2 (en) | 2004-03-04 | 2012-08-28 | Media Rights Technologies, Inc. | Method for providing curriculum enhancement using a computer-based media access system |
WO2005088507A1 (en) | 2004-03-04 | 2005-09-22 | Yates James M | Method and apparatus for digital copyright exchange |
US20060080703A1 (en) | 2004-03-22 | 2006-04-13 | Compton Charles L | Content storage method and system |
US20050222900A1 (en) | 2004-03-30 | 2005-10-06 | Prashant Fuloria | Selectively delivering advertisements based at least in part on trademark issues |
NZ551233A (en) | 2004-04-14 | 2009-04-30 | Digital River Inc | Geographic location based licensing system |
US8688248B2 (en) | 2004-04-19 | 2014-04-01 | Shazam Investments Limited | Method and system for content sampling and identification |
US20050276570A1 (en) | 2004-06-15 | 2005-12-15 | Reed Ogden C Jr | Systems, processes and apparatus for creating, processing and interacting with audiobooks and other media |
US8953908B2 (en) | 2004-06-22 | 2015-02-10 | Digimarc Corporation | Metadata management and generation using perceptual features |
JP2006039791A (en) | 2004-07-26 | 2006-02-09 | Matsushita Electric Ind Co Ltd | Transmission history dependent processor |
US7631336B2 (en) | 2004-07-30 | 2009-12-08 | Broadband Itv, Inc. | Method for converting, navigating and displaying video content uploaded from the internet to a digital TV video-on-demand platform |
JP4817624B2 (en) | 2004-08-06 | 2011-11-16 | キヤノン株式会社 | Image processing system, image alteration judgment method, computer program, and computer-readable storage medium |
US7860922B2 (en) | 2004-08-18 | 2010-12-28 | Time Warner, Inc. | Method and device for the wireless exchange of media content between mobile devices based on content preferences |
US7555487B2 (en) | 2004-08-20 | 2009-06-30 | Xweb, Inc. | Image processing and identification system, method and apparatus |
US20060058019A1 (en) | 2004-09-15 | 2006-03-16 | Chan Wesley T | Method and system for dynamically modifying the appearance of browser screens on a client device |
US8117339B2 (en) | 2004-10-29 | 2012-02-14 | Go Daddy Operating Company, LLC | Tracking domain name related reputation |
GB0424479D0 (en) | 2004-11-05 | 2004-12-08 | Ibm | Generating a fingerprint for a document |
US20060128470A1 (en) | 2004-12-15 | 2006-06-15 | Daniel Willis | System and method for managing advertising content delivery in a gaming environment supporting aggregated demographics serving and reporting |
US7562228B2 (en) | 2005-03-15 | 2009-07-14 | Microsoft Corporation | Forensic for fingerprint detection in multimedia |
US20070242880A1 (en) | 2005-05-18 | 2007-10-18 | Stebbings David W | System and method for the identification of motional media of widely varying picture content |
US8365306B2 (en) | 2005-05-25 | 2013-01-29 | Oracle International Corporation | Platform and service for management and multi-channel delivery of multi-types of contents |
US20060287916A1 (en) | 2005-06-15 | 2006-12-21 | Steven Starr | Media marketplaces |
US20060287996A1 (en) | 2005-06-16 | 2006-12-21 | International Business Machines Corporation | Computer-implemented method, system, and program product for tracking content |
US7788132B2 (en) | 2005-06-29 | 2010-08-31 | Google, Inc. | Reviewing the suitability of Websites for participation in an advertising network |
JP2007011554A (en) | 2005-06-29 | 2007-01-18 | Konica Minolta Business Technologies Inc | Image forming apparatus |
US7925973B2 (en) | 2005-08-12 | 2011-04-12 | Brightcove, Inc. | Distribution of content |
WO2007052171A2 (en) | 2005-09-01 | 2007-05-10 | Zvi Haim Lev | System and method for reliable content access using a cellular/wireless device with imaging capabilities |
US7697942B2 (en) | 2005-09-02 | 2010-04-13 | Stevens Gilman R | Location based rules architecture systems and methods |
US20070106551A1 (en) | 2005-09-20 | 2007-05-10 | Mcgucken Elliot | 22nets: method, system, and apparatus for building content and talent marketplaces and archives based on a social network |
US20080004116A1 (en) | 2006-06-30 | 2008-01-03 | Andrew Stephen Van Luchene | Video Game Environment |
US20070162349A1 (en) | 2005-10-17 | 2007-07-12 | Markmonitor Inc. | Client Side Brand Protection |
JP4629555B2 (en) | 2005-11-07 | 2011-02-09 | インターナショナル・ビジネス・マシーンズ・コーポレーション | Restoration device, program, information system, restoration method, storage device, storage system, and storage method |
US7412224B2 (en) | 2005-11-14 | 2008-08-12 | Nokia Corporation | Portable local server with context sensing |
KR20080073357A (en) | 2005-11-29 | 2008-08-08 | 구글 인코포레이티드 | Detecting repeating content in broadcast media |
JP4154421B2 (en) | 2005-12-07 | 2008-09-24 | キヤノン株式会社 | Image processing apparatus, program for executing the image processing method, and medium storing the program |
WO2007076459A2 (en) | 2005-12-21 | 2007-07-05 | Digimarc Corporation | Rules driven pan id metadata routing system and network |
US20070162761A1 (en) | 2005-12-23 | 2007-07-12 | Davis Bruce L | Methods and Systems to Help Detect Identity Fraud |
US20070157228A1 (en) | 2005-12-30 | 2007-07-05 | Jason Bayer | Advertising with video ad creatives |
US20070260520A1 (en) | 2006-01-18 | 2007-11-08 | Teracent Corporation | System, method and computer program product for selecting internet-based advertising |
WO2007089943A2 (en) | 2006-02-01 | 2007-08-09 | Markmonitor Inc. | Detecting online abuse in images |
US8122019B2 (en) | 2006-02-17 | 2012-02-21 | Google Inc. | Sharing user distributed search results |
US20080027931A1 (en) | 2006-02-27 | 2008-01-31 | Vobile, Inc. | Systems and methods for publishing, searching, retrieving and binding metadata for a digital object |
US20070203891A1 (en) | 2006-02-28 | 2007-08-30 | Microsoft Corporation | Providing and using search index enabling searching based on a targeted content of documents |
US20070208715A1 (en) | 2006-03-02 | 2007-09-06 | Thomas Muehlbauer | Assigning Unique Content Identifiers to Digital Media Content |
US8037506B2 (en) | 2006-03-03 | 2011-10-11 | Verimatrix, Inc. | Movie studio-based network distribution system and method |
US20070233556A1 (en) | 2006-03-31 | 2007-10-04 | Ross Koningstein | Controlling the serving, with a primary document, of ads from a first source, subject to a first compensation scheme, and ads from a second source, subject to a second compensation scheme |
US8009861B2 (en) | 2006-04-28 | 2011-08-30 | Vobile, Inc. | Method and system for fingerprinting digital video object based on multiresolution, multirate spatial and temporal signatures |
US8745226B2 (en) | 2006-05-02 | 2014-06-03 | Google Inc. | Customization of content and advertisements in publications |
US7831531B1 (en) | 2006-06-22 | 2010-11-09 | Google Inc. | Approximate hashing functions for finding similar content |
MY166373A (en) | 2006-06-23 | 2018-06-25 | Tencent Tech Shenzhen Co Ltd | Method, system and apparatus for playing advertisements |
US20080051029A1 (en) | 2006-08-25 | 2008-02-28 | Bradley James Witteman | Phone-based broadcast audio identification |
US8707459B2 (en) | 2007-01-19 | 2014-04-22 | Digimarc Corporation | Determination of originality of content |
US20080059211A1 (en) | 2006-08-29 | 2008-03-06 | Attributor Corporation | Content monitoring and compliance |
US8010511B2 (en) | 2006-08-29 | 2011-08-30 | Attributor Corporation | Content monitoring and compliance enforcement |
US20080059285A1 (en) | 2006-09-01 | 2008-03-06 | Admob, Inc. | Assessing a fee for an ad |
JP5546246B2 (en) * | 2006-11-03 | 2014-07-09 | グーグル インコーポレイテッド | Content management system |
US7707224B2 (en) | 2006-11-03 | 2010-04-27 | Google Inc. | Blocking of unlicensed audio content in video files on a video hosting website |
US20080155701A1 (en) | 2006-12-22 | 2008-06-26 | Yahoo! Inc. | Method and system for unauthorized content detection and reporting |
US20080162449A1 (en) | 2006-12-28 | 2008-07-03 | Chen Chao-Yu | Dynamic page similarity measurement |
US7979464B2 (en) | 2007-02-27 | 2011-07-12 | Motion Picture Laboratories, Inc. | Associating rights to multimedia content |
US20080235795A1 (en) | 2007-03-23 | 2008-09-25 | Ishikawa Mark M | System and Method for Confirming Digital Content |
JP5337147B2 (en) | 2007-05-03 | 2013-11-06 | グーグル インコーポレイテッド | Converting digital content postings into cash |
US8117094B2 (en) | 2007-06-29 | 2012-02-14 | Microsoft Corporation | Distribution channels and monetizing |
US8238669B2 (en) * | 2007-08-22 | 2012-08-07 | Google Inc. | Detection and classification of matches between time-based media |
US8490206B1 (en) | 2007-09-28 | 2013-07-16 | Time Warner, Inc. | Apparatuses, methods and systems for reputation/content tracking and management |
-
2012
- 2012-11-27 US US13/686,541 patent/US10242415B2/en not_active Expired - Fee Related
-
2019
- 2019-03-21 US US16/360,896 patent/US20200058088A1/en not_active Abandoned
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20220027407A1 (en) * | 2020-07-27 | 2022-01-27 | Audible Magic Corporation | Dynamic identification of unknown media |
Also Published As
Publication number | Publication date |
---|---|
US10242415B2 (en) | 2019-03-26 |
US20130085825A1 (en) | 2013-04-04 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US20200058088A1 (en) | Method and system for determining content treatment | |
US9785841B2 (en) | Method and system for audio-video signal processing | |
US10007723B2 (en) | Methods for identifying audio or video content | |
US9396312B2 (en) | Syndication including melody recognition and opt out | |
US20230075884A1 (en) | Systems and Methods for Token Management in Social Media Environments | |
US9098807B1 (en) | Video content claiming classifier | |
Fleischer et al. | Discovering Spotify-A thematic introduction | |
US20150237410A1 (en) | System and method for providing content-aware persistent advertisements | |
US20140040088A1 (en) | Media rights management using melody identification | |
CN102216945B (en) | Networking with media fingerprints | |
US20230011621A1 (en) | Artifact Origination and Content Tokenization | |
US11935094B2 (en) | Managing metadata enrichment of digital asset portfolios | |
KR20200044692A (en) | Display apparatus and the control method thereof for advertisement content transmission | |
Berkowitz | The Treble with Technocracy: Reconsidering Patron Interactions With Streaming Music | |
Savoia | Streaming technology and its impact on the music industry A logistic regression analysis of willingness to pay for digital music and premium subscriptions to streaming music platforms | |
Thomas | A Consumer Investigation: Monetising, Marketing and Distributing Digital Music in a Niche and Minority Language Market–The Situation in Wales | |
Krohn | Media-Rich Input Application Liability | |
KR20080078748A (en) | Contents service system and method | |
Grassmuck | IFPI and Economists on File-Sharing and the Music Industry A Literature Review |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: DIGIMARC CORPORATION, OREGON Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:DAVIS, BRUCE L.;CONWELL, WILLIAM Y.;SIGNING DATES FROM 20190904 TO 20190905;REEL/FRAME:050280/0042 |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: APPLICATION DISPATCHED FROM PREEXAM, NOT YET DOCKETED |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: NON FINAL ACTION MAILED |
|
STCB | Information on status: application discontinuation |
Free format text: ABANDONED -- FAILURE TO RESPOND TO AN OFFICE ACTION |