CN101601280A - Be used to make the method and apparatus of the transitions smooth between first video-frequency band and second video-frequency band - Google Patents
Be used to make the method and apparatus of the transitions smooth between first video-frequency band and second video-frequency band Download PDFInfo
- Publication number
- CN101601280A CN101601280A CNA2008800039574A CN200880003957A CN101601280A CN 101601280 A CN101601280 A CN 101601280A CN A2008800039574 A CNA2008800039574 A CN A2008800039574A CN 200880003957 A CN200880003957 A CN 200880003957A CN 101601280 A CN101601280 A CN 101601280A
- Authority
- CN
- China
- Prior art keywords
- video
- frequency band
- profile
- feature
- definite
- 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.)
- Pending
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N5/00—Details of television systems
- H04N5/222—Studio circuitry; Studio devices; Studio equipment
- H04N5/262—Studio circuits, e.g. for mixing, switching-over, change of character of image, other special effects ; Cameras specially adapted for the electronic generation of special effects
- H04N5/268—Signal distribution or switching
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Television Signal Processing For Recording (AREA)
- Management Or Editing Of Information On Record Carriers (AREA)
- Apparatus For Radiation Diagnosis (AREA)
- Image Processing (AREA)
- Picture Signal Circuits (AREA)
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
Abstract
Make the first video-frequency band (C by following operation
1..., C
M) and the second video-frequency band (S
1..., S
N) between transitions smooth: determine the first content profile of (103) first video-frequency bands, determine the second content profile of (103) second video-frequency bands subsequently; And at the position (S of determined first profile similar in appearance to second profile
j, S
J+1) locate first video-frequency band is inserted in second video-frequency band so that the transitions smooth between first video-frequency band and second video-frequency band.
Description
Technical field
The present invention relates to be used to make the method and apparatus of the transitions smooth between first video-frequency band and second video-frequency band.
Background technology
Because developing rapidly of DMB and distribution, commercial advertisement now just plays an important role in daily life.There is not under the situation of commercial advertisement the broadcast receiver of acquisition or TV program become rare day by day.It is that its product is advertised that company uses commercial advertisement, and broadcaster needs commercial advertisement to support (perhaps even main) income.What on the other hand, ordinary consumer was considered as commercial insert that unsolicited ground invades them usually watches or listens to experience.Therefore, the consumer uses video tape recorder to skip these commercial-block, and this has reduced the advertising income of broadcaster.
Summary of the invention
Desirablely be, make commercial insert (or piece) thus transitions smooth become not too important or not too make us expectation to improve the viewing experience commercial skip.
This can be according to a first aspect of the invention by being used to make the method for the transitions smooth between first video-frequency band and second video-frequency band realize that effectively this method may further comprise the steps: the first content profile (profile) of determining first video-frequency band; Determine the second content profile of second video-frequency band; And first video-frequency band is inserted in second video-frequency band so that the transitions smooth between first video-frequency band and second video-frequency band in the position of determined first profile similar in appearance to determined second profile.
This also can be according to a second aspect of the invention by being used to make the equipment of the transitions smooth between first video-frequency band and second video-frequency band realize that this equipment comprises: be used for determining first video-frequency band the first content profile first determine device; The second definite device that is used for the second content profile of definite second video-frequency band; And being used for determining that first video-frequency band is inserted into the 3rd definite device of the position in second video-frequency band, first profile of determining in this position is similar in appearance to second profile of determining, so that the transitions smooth between first video-frequency band and second video-frequency band.
In this way, provide one group of simple video editing option, thereby realized more seamless integrated of given commercial advertisement and other non-commercial content (for example, the content of narration is such as the movie or television play).This editor is intended to preserve essential information, the sudden change of the transition that obtains commercial insert or piece is minimized and make not think commercial skip so.The present invention is effective especially when the film that is used for specialty and TV broadcasting and editor.
Effectively, system of the present invention has changed the commercial advertisement INS effectively.In addition, each commercial advertisement in this piece can be rearranged, and can further be modified in the audio-visual content of the boundary between described each commercial advertisement, and from/revise audio-visual content during to the transition of the non-commercial content of adjacency.When commercial advertisement is inserted in the fixed position, for example determined to insert the fixed time of commercial advertisement when creator of content, also can be modified between each commercial advertisement boundary and from/to the audio-visual content of the transition position of the non-commercial content of adjacency.
In addition, in a preferred embodiment, can verify this system, promptly detect editor's commercial advertisement once by using method and strategy, and if commercial advertisement remain detectable, then by its feedback (by this system) is edited described material.
Summarize the content of commercial insert, and can insert the selection of the position of commercial advertisement according to this profile.In fact, can restricted (for example only after scene finish) for the specified point that inserts commercial advertisement, and have demonstration (for example enter content 15 to 20 minutes between) for the general position in the content.In these restrictions, the optimum position that can select to insert commercial advertisement is so that the difference between commercial advertisement and the sealing content minimizes, and so provides desirable and seamlessly transit.
And when summarizing the content of commercial insert, each commercial advertisement in this piece can be rearranged.Determine selection about order according to profile separately, commercial advertisement should be adjacent in proper order and places towards the border with non-commercial content with this.This can be used to make the common big audiovisual of the commercial-block inside cunning that flattens.In fact, it is the pattern of the frequent and unexpected interruption of a plurality of audiovisual features in the relatively short time cycle (a few minutes), and this has destructive and horrible for the observer especially.
The audio-visual content of the boundary between each commercial advertisement can be modified, and can be modified in from/to the audio-visual content of the transition position of adjacent non-commercial content.Be well known that, progressively transition between vision (camera) camera lens (for example cross fade and fade out) not too has destructiveness and more is difficult to detect than unexpected cut-out, for example as Ying Li, the article of C.-C.Jay Kuo " VideoContent Analysis Using Multimodal Information " (2003, KluwerAcademic Publishers Group, ISBN 1-4020-7490-5) disclosed in.Therefore, when the boundary between each commercial advertisement provides so progressively transition, will further make transitions smooth and disturb the shear of two-forty vision lens effectively or the detection of the super separator of audiovisual.Also can create similar effect in audio frequency, it also can be useful wherein inserting non-audible noise.
Description of drawings
In order more intactly to understand the present invention, reference is below in conjunction with the explanation that accompanying drawing carried out, in the accompanying drawings:
Fig. 1 is the rough schematic view of equipment according to an embodiment of the invention;
Fig. 2 (a)-(d) illustrates first example of the low-level features statistics in the feature film (feature movie) that comprises commercial-block; And
Fig. 3 (a)-(d) illustrates second example of the low-level features statistics in the feature film that comprises commercial-block;
Embodiment
With reference to Fig. 1, with the equipment of embodiment of the present invention will be described in more detail.Equipment 100 comprises the input terminal 101 that is used for receiving multimedia data stream.Input terminal 101 is connected to the input of multimedia (audiovisual) content analyser 103 and the input of video editor 105.The output of video editor 105 is connected to the lead-out terminal 107 of equipment 100.The output of content of multimedia analyzer 103 is connected to the controller 109 of video editor 105.The output of video editor 105 also is connected to the controller 109 of video editor 105 through reference detector 111 in feedback loop.
Will be discussed in more detail below the operation of described equipment now.
Comprise the second video-frequency band S
1..., S
NWith the first video-frequency band C
1..., C
MMultimedia data stream be imported on the input terminal 101 of equipment 100.For simplicity, first and second video-frequency bands are shown as input on input terminal 101 individually.First video-frequency band is for example by a plurality of independently commercial advertisement C
1..., C
MOr information-based data segment is formed.Second video-frequency band is by continuous sequence (set) S as vision lens
1..., S
NNoncommercial advertising video content broadcasting form.First and/or second video-frequency band not only comprises the data of vision but also comprise the corresponding audio data.
This input multimedia data stream is analyzed by content of multimedia analyzer 103.At first for example discern the first and second video-frequency band C according to the audiovisual inspection of being undertaken by the people
1..., C
MAnd S
1..., S
N, and be easier index possibly it is made marks with visit (by means of relevant metadata).Subsequently, extract the feature of the performance characteristics that can be used as first and second video-frequency bands.For the technical staff, the method that a large amount of audiovisual features and being used to is extracted these features is known.List some illustrative example now:
Advanced features.
The existing of people (can pass through face or speech recognition identification) in-the scene
-object or speaker follow the tracks of (comprising the detection that the speaker changes)
The tone of-content (for example obtaining) from the tone of music or by the analyzing speech rhythm
Mid-level features
The audio frequency composition of-scene (sound, music, sound+music, sound+background noise, or the like)
The location of-vision (camera) shot-cut.
The detection that-audiovisual delimiter exists, promptly super separator (monochrome video frame and noiseless the combining of audio frequency)
The detection that-veneer text (overlaid text), broadcaster's sign or the like exist.
Low-level features
-vision:
-mass-tone (for example color of maximum group in color space)
-brightness (luma), pigment (color) are average, histogram, gradient or the like
The level of-visual activity and gradient (for example the statistics from coding parameter (for example motion vector) obtains)
The level of-scene complexity and gradient (for example the statistics from coding parameter (such as the product of encoded bit rate and quantization parameter) obtains)
-audio frequency (time and spectral property):
-volume;
-(for example speaker) speed;
-background noise characteristic;
-(for example speaker) tone dynamic characteristic.
Handle the feature of being extracted by analyzer 103 subsequently and come control of video editing machine 105 to generate the content profile.
The content profile is the estimation according to the content similarity of the feature of being extracted.These profiles can generate by following diverse ways.
Profile can typically be made up of characteristic statistics, and characteristic statistics is for example gone up average and the standard deviation of calculating at each feature in a plurality of continuous video frames (analysis window).For advanced features, standard deviation may be the most meaningful, and other other measurements that are suitable for binary signal also it is contemplated that.
In first embodiment, each candidate feature is considered separately, and is combined to form final sumbission from the result that different features obtain.Therefore, at the single feature contour of content creating of first video-frequency band and second video-frequency band.Relatively these single feature contours are estimated (confidence level, probability) to produce similarity.Can obtain this estimation by for example measuring tolerance (metric), described tolerance is big more such as the more little similarity of distance-distance.Use technique known that these a plurality of estimations are combined as single decision-making subsequently, technique known is such as majority voting, the linear decision model with weighting, fuzzy logic, Markov (Markov) model or the like.
In alternate embodiments, from the combination of low-level features, obtain the multidimensional characteristic vectors that synthetic profile-acquisition comprises (as above-mentioned) statistics of each feature.Subsequently, use the similarity between the such characteristic vector of extracting from different content items of the commercial measurement know from the statistical pattern classification field, for example described technology is by calendar year 2001 John Wiley﹠amp; Sons, the Richard O.Duda among the ISBN 0-471-05669-3, Peter E.Hart, " the Pattern Classification " of David G.Stork is disclosed, for example data clusters.This can realize by using the technology such as supervised learning and neural net.
At last, should be noted that the combination advanced features may be inappropriate or or even infeasible (for example quantizing advanced features may be difficult such as the speaker's tracking or the content tone) with the measurement of determining similarity in multidimensional characteristic vectors.In this case, for example assess feature independently to obtain similarity measurement, can use above-mentioned technology to combine them after this by using heuritic approach.
Further explain the notion of summarizing according to the content of the foregoing description, above-mentioned two examples that illustrate the statistics of such feature with reference to Fig. 2 (a)-(d) and 3 (a)-(d).In example, show feature film and animation cartoon sequence.
Fig. 2 (a) and 3 (a) expression True Data (artificial mark), wherein 1 corresponding to commercial advertisement, and 0 corresponding to non-commercial content.Fig. 2 (b) and 3 (b) are expressed as data the standard deviation of mean flow rate of a frame of video of this example.In each frame position, this can calculate on the analysis window that is positioned at 3500 frame of video at this place-centric place (～2.5 minutes PAL videos).Fig. 2 (c) and 3 (c) are the probability of the voice of the same sample of presentation graphs 2 (b) and 3 (b) respectively.Fig. 2 (d) and 3 (d) are the probability of the music of the same sample of presentation graphs 2 (b), 2 (c), 3 (b) and 3 (c) respectively.
In order to illustrate, by factor 2 with the initial data sub sampling.
In first example of Fig. 2 (a)-(d), the most distinguishing feature is the speech probability shown in Fig. 2 (c).To present commercial-block is separated, each commercial-block is obviously being created individual " engraving plateau " in the low amplitude cinematic data.
When considering commercial-block CB2, if thereby having inserted identical commercial-block after the little meeting youngster makes that it will be near the peak value subsequently of non-commercial content, then this will create " plateau " that covers a commercial-block and a film simultaneously.Therefore, this will utilize cinematic data subsequently to create more seamless transition.This transition should mainly be that the automatic categorizer and the people of the feature of commercial-block presents seamless for understanding " plateau ".For CB3, can similarly observe, if be inserted into subsequently, then will utilize cinematic data subsequently to create more seamless transition.Referring again to Fig. 2 (c).
In second example of Fig. 3 (a)-(d), mean flow rate provides the best clue of separating, and audio frequency characteristics is very as broad as long.In this case, can observe additional difference between them in commercial-block.The 3rd commercial-block is created the most significant " plateau ", and it can be distinguished mutually with the corresponding plateau of the first and the 4th commercial-block, and this can clearly observe from Fig. 3 (b).Can be by each commercial advertisement be redistributed in different pieces to realize the leveling on this plateau.
Imaginabalely be, for some other styles, these features all will not have difference, and some other features then are not.Exactly, the ability to see things in their true light of low-level features makes a plurality of features of combination become preferably so that generate profile to this dependence of content type (style).This also is favourable, because any interference that editor produces in a feature output may be " multiplication ", promptly appears in other features.Even can imagine, artificial mode may appear in some common indiscriminate features of the effect that increases video editor.
The output of above-mentioned analysis is imported in the controller 109 of video editor 105 being used for and recommends certain editor's action or automatically perform suitable editor's action to user (broadcaster).
Such editor's possible outcome has been shown among Fig. 1.Commercial-block C
3C
1Be combined and be inserted in lens group S
JWith S
J+1Between.This may be because S
JDwell section be found similar in appearance to C
3The beginning part, and S
J+1Beginning part similar in appearance to C
1Dwell section.Otherwise be because from C
3To C
1Transition position observe high similarity.Therefore, can observe level and smooth transition between noncommercial advertising part and commercial advertisement, this makes and the more pleasant automatic commercial-block detection of watching and helping to stop PVR occurs.In addition, can insert the section of being made up of extra content " T ", owing to noiseless insertion well known in the art, cross compound turbine etc., described extra content can produce between different sections.
Can also use compressed video (promptly after coding) executive editor's operation, this compressed video is common in the video production of specialty.And, can imagine that reference detector 111 and multimedia (audiovisual) content analyser 103 can be overlapping, because they can comprise a plurality of identical operations.
Data flow output from the editor of editing machine 105 is fed back to the controller 109 of editing machine 105 to adjust editing machines 105 by reference detector 111.Reference detector 111 comprises known commercial block detector, and it seeks the transition between noncommercial advertising part and the commercial-block so that distinguish commercial advertisement and noncommercial advertising part.If unsmooth by the transition that editing machine 105 is created, thereby then this will be detected and be fed to controller 109 by reference block 111 and improve level and smooth to the transition between the different video section with the operation of adjusting editing machine 105.The data flow of being edited is placed on the lead-out terminal 107 of equipment 100 subsequently.
Although described the present invention in conjunction with the preferred embodiments, but be to be understood that, therefore to those skilled in the art, the change to described embodiment in above-mentioned principle will be obvious, and the invention is not restricted to these preferred embodiments and be intended to comprise this change.The invention reside in each combination of each novel property feature and these property features.Reference numeral in the claim is its scope of protecting without limits.Verb " comprises " and the existence of other elements outside narrating in the claim is not got rid of in the use of being out of shape.The existence of a plurality of such elements is not got rid of in the use of article before element " ".
As for a person skilled in the art clearly, the meaning of " device " is hardware (such as discrete or integrated circuit or electronic component) or the software (such as the part of program or program) that comprises any execution in operation or be designed to carry out specific function, no matter still be individually, also still work collaboratively no matter isolatedly with other elements in conjunction with other functions.The present invention can be by comprising element that several are different and realizing by the computer of suitable programming.In listing the equipment claim of several means, the several of these devices can be realized by same hardware." computer program " be understood that to represent to be stored on the computer-readable medium, can be through network download or any software product that can sell by any way.
Claims (11)
1. method that makes the transitions smooth between first video-frequency band and second video-frequency band, this method may further comprise the steps:
-determine the first content profile of first video-frequency band;
-determine the second content profile of second video-frequency band; And
-described first video-frequency band is inserted in described second video-frequency band so that the transitions smooth between described first video-frequency band and described second video-frequency band in the position of determined first profile similar in appearance to determined second profile.
2. according to the process of claim 1 wherein that the step that described first video-frequency band is inserted in described second video-frequency band comprises:
The position that-similarity between determined first profile and determined second profile surpasses similarity threshold is inserted into described first video-frequency band in the described second definite video-frequency band.
3. according to the process of claim 1 wherein that the step of determining first profile may further comprise the steps:
-each frame from a plurality of frames of described first video-frequency band extracts at least one first feature;
-determine the single order or the statistical attribute of high-order more of first feature of described extraction; And
First profile of described definite statistical attribute of first feature of the described extraction of-generation.
4. according to the method for claim 3, determine that wherein the step of second profile may further comprise the steps:
-each frame from a plurality of frames of described second video-frequency band extracts at least one second feature, and described at least one second feature is corresponding to described at least one first feature;
-determine the single order or the statistical attribute of high-order more of second feature of described extraction; And
Second profile of described definite statistical attribute of second feature of the described extraction of-generation.
5. according to the process of claim 1 wherein that the step that described first video-frequency band is inserted in described second video-frequency band may further comprise the steps:
-basis is determined a plurality of similarities estimations at first profile of the described generation of first and second features of a plurality of extractions and second profile of described generation;
-described a plurality of similarities are estimated combination;
-estimate that in the similarity of described combination the place that is higher than predetermined threshold is inserted into described first video-frequency band in described second video-frequency band.
6. according to the process of claim 1 wherein that the step that above-mentioned first video-frequency band is inserted in described second video-frequency band may further comprise the steps:
-basis is determined a plurality of similarities estimations at first profile of the described generation of the part of a plurality of correspondences of first and second features and second profile of described generation;
-determine that the highest similarity that described a plurality of similarity is estimated estimates;
-in the position that described highest similarity is estimated described first video-frequency band is inserted in described second video-frequency band.
7. according to the process of claim 1 wherein that this method further comprises:
According to any difference that stays between described first profile and described second profile insertion portion is inserted into transition position between described first video-frequency band and described second video-frequency band, so that make the transitions smooth between described first video-frequency band and described second video-frequency band.
8. computer program comprises a plurality of program code parts that are used for realizing according to any one method of aforementioned claim.
9. equipment that makes the transitions smooth between first video-frequency band and second video-frequency band, this equipment comprises:
-be used for determining first video-frequency band the first content profile first determine device;
-be used for determining second video-frequency band the second content profile second determine device; And
-be used for determining that first video-frequency band is inserted into the 3rd of position in second video-frequency band determines device, the described first definite profile is similar in appearance to the described second definite profile, so that the transitions smooth between described first video-frequency band and described second video-frequency band in this position.
10. according to the equipment of claim 9, wherein this equipment further comprises:
Be used for first video-frequency band being inserted into the editing device in described second video-frequency band in described definite position.
11. according to the equipment of claim 9, wherein this equipment further comprises:
-extraction element be used for extracting at least one first feature from each frame of a plurality of frames of described first video-frequency band, and/or each frame from a plurality of frames of described second video-frequency band extracts at least one second feature; And
-processing unit, be used for determining described extraction first feature single order or more the higher order statistical attribute and generate described extraction first feature described definite statistical attribute first profile and/or determine the single order or the higher order statistical attribute and generate second profile of described definite statistical attribute of described second feature more of second feature of described extraction.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| EP07101558.0 | 2007-02-01 | ||
| EP07101558 | 2007-02-01 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101601280A true CN101601280A (en) | 2009-12-09 |
Family
ID=39563508
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2008800039574A Pending CN101601280A (en) | 2007-02-01 | 2008-01-28 | Be used to make the method and apparatus of the transitions smooth between first video-frequency band and second video-frequency band |
Country Status (3)
| Country | Link |
|---|---|
| JP (1) | JP2010518672A (en) |
| CN (1) | CN101601280A (en) |
| WO (1) | WO2008093277A2 (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114946192A (en) * | 2020-01-15 | 2022-08-26 | 杜比国际公司 | Adaptive streaming media content with bit rate switching |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5924127B2 (en) * | 2012-05-24 | 2016-05-25 | カシオ計算機株式会社 | Movie generation apparatus, movie generation method, and program |
| WO2018007951A1 (en) * | 2016-07-07 | 2018-01-11 | Corephotonics Ltd. | Dual-camera system with improved video smooth transition by image blending |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2001030073A1 (en) * | 1999-10-19 | 2001-04-26 | Koninklijke Philips Electronics N.V. | Television receiver and method of using same for displaying information messages |
-
2008
- 2008-01-28 CN CNA2008800039574A patent/CN101601280A/en active Pending
- 2008-01-28 JP JP2009547788A patent/JP2010518672A/en not_active Withdrawn
- 2008-01-28 WO PCT/IB2008/050296 patent/WO2008093277A2/en active Application Filing
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114946192A (en) * | 2020-01-15 | 2022-08-26 | 杜比国际公司 | Adaptive streaming media content with bit rate switching |
| US11997339B2 (en) | 2020-01-15 | 2024-05-28 | Dolby International Ab | Adaptive streaming of media content with bitrate switching |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2008093277A2 (en) | 2008-08-07 |
| WO2008093277A3 (en) | 2008-10-23 |
| JP2010518672A (en) | 2010-05-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US6928233B1 (en) | Signal processing method and video signal processor for detecting and analyzing a pattern reflecting the semantics of the content of a signal | |
| US7327885B2 (en) | Method for detecting short term unusual events in videos | |
| Brezeale et al. | Automatic video classification: A survey of the literature | |
| CN111683209A (en) | Mixed-cut video generation method and device, electronic equipment and computer-readable storage medium | |
| US6744922B1 (en) | Signal processing method and video/voice processing device | |
| CN110119711A (en) | A kind of method, apparatus and electronic equipment obtaining video data personage segment | |
| JP5686800B2 (en) | Method and apparatus for processing video | |
| US20060114992A1 (en) | AV signal processing apparatus for detecting a boundary between scenes, method, recording medium and computer program therefor | |
| KR20000054561A (en) | A network-based video data retrieving system using a video indexing formula and operating method thereof | |
| US6996171B1 (en) | Data describing method and data processor | |
| JP5360979B2 (en) | Important information extraction method and apparatus | |
| Mizutani et al. | Commercial detection in heterogeneous video streams using fused multi-modal and temporal features | |
| Sanchez et al. | Shot partitioning based recognition of tv commercials | |
| EP1569240A1 (en) | Apparatus and method for determining anchor shots | |
| Gagnon et al. | Towards computer-vision software tools to increase production and accessibility of video description for people with vision loss | |
| CN101601280A (en) | Be used to make the method and apparatus of the transitions smooth between first video-frequency band and second video-frequency band | |
| US20060074893A1 (en) | Unit for and method of detection a content property in a sequence of video images | |
| JP5257356B2 (en) | Content division position determination device, content viewing control device, and program | |
| JP2009060413A (en) | Method and system for extracting feature of moving image, and method and system for retrieving moving image | |
| Benini et al. | Hidden Markov models for video skim generation | |
| Zhu et al. | SVM-based video scene classification and segmentation | |
| Chen et al. | An effective method for video genre classification | |
| Bailer et al. | Skimming rushes video using retake detection | |
| WO2024034401A1 (en) | Video editing device, video editing program, and video editing method | |
| El-Khoury et al. | Unsupervised TV program boundaries detection based on audiovisual features |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Open date: 20091209 |