CN110019852A - Multimedia resource searching method and device - Google Patents
Multimedia resource searching method and device Download PDFInfo
- Publication number
- CN110019852A CN110019852A CN201711444505.7A CN201711444505A CN110019852A CN 110019852 A CN110019852 A CN 110019852A CN 201711444505 A CN201711444505 A CN 201711444505A CN 110019852 A CN110019852 A CN 110019852A
- Authority
- CN
- China
- Prior art keywords
- multimedia resource
- phrases
- key message
- words
- information
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/40—Information retrieval; Database structures therefor; File system structures therefor of multimedia data, e.g. slideshows comprising image and additional audio data
- G06F16/43—Querying
- G06F16/432—Query formulation
- G06F16/433—Query formulation using audio data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/40—Information retrieval; Database structures therefor; File system structures therefor of multimedia data, e.g. slideshows comprising image and additional audio data
- G06F16/43—Querying
- G06F16/435—Filtering based on additional data, e.g. user or group profiles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/40—Information retrieval; Database structures therefor; File system structures therefor of multimedia data, e.g. slideshows comprising image and additional audio data
- G06F16/48—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/40—Information retrieval; Database structures therefor; File system structures therefor of multimedia data, e.g. slideshows comprising image and additional audio data
- G06F16/48—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/487—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using geographical or spatial information, e.g. location
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Library & Information Science (AREA)
- Mathematical Physics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
This disclosure relates to a kind of multimedia resource searching method and device.Wherein this method includes carrying out semantic analysis to search words and phrases, obtains each association words and phrases relevant to search words and phrases；Search the information labels of each multimedia resource related with each association sentence, the information labels of the multimedia resource include the text information of natural language associated with the key message of multimedia resource, and the key message of the multimedia resource is to carry out feature extraction to the content of multimedia resource to obtain；According to the information labels of each multimedia resource, multimedia resource included by search result is determined.By the way that the associate text information of the content of multimedia resource and Human Natural Language is got up, when search, semantic analysis is carried out to search words and phrases again and obtains association sentence, the information labels of corresponding multimedia resource can be searched, and then be more accurately obtained the content of required multimedia resource.
Description
Technical field
This disclosure relates to multimedia technology field more particularly to a kind of multimedia resource searching method and device.
Background technique
During video content creation, theme is needed and created, the video element that the expression wish of creator is consistent
Material.But the material of video, audio is difficult to find suitable material by way of text.And current search engine, all it is
What text included in the structural description data based on the page or movie and television play that link together with video carried out, Wu Fajing
Standard navigates to the material that creator wants.
The search engine of mainstream, video website etc. at present are provided based on text informations such as title, description, related names
Carry out the service of video search.But these text informations are based on, the specific requirements of creator are unable to satisfy.For example, short-sighted frequency
Creator generally can organize a series of story boards, picture, and suitable audio to be combined according to the theme being intended by.
Currently there are search engine do not have search capability required for video creator.In other words, for vertical neck
The search in domain, universal search engine are difficult the solution having had.
Summary of the invention
In view of this, the present disclosure proposes a kind of multimedia resource searching method and devices.
According to the one side of the disclosure, a kind of multimedia resource searching method is provided, comprising:
Semantic analysis is carried out to search words and phrases, obtains each association words and phrases relevant to search words and phrases；
Search the information labels of each multimedia resource related with each association sentence, the information labels of the multimedia resource
Text information including natural language associated with the key message of multimedia resource, the key message of the multimedia resource
It is to carry out feature extraction to the content of multimedia resource to obtain；
According to the information labels of each multimedia resource, multimedia resource included by search result is determined.
In one possible implementation, semantic analysis is carried out to search words and phrases, obtained relevant to search words and phrases each
It is associated with words and phrases, comprising:
Semantic analysis is carried out to search words and phrases, is obtained and the search relevant each retrieval words and phrases of words and phrases and each filtering words and phrases.
In one possible implementation, the information labels of each multimedia resource related with each association sentence are searched,
It include: to search each multimedia resource related with each retrieval words and phrases in the key message association lexicon of multimedia resource
First information label；
According to the information labels of each multimedia resource, multimedia resource included by search result is determined, comprising: by first
Multimedia resource corresponding to information labels, as the first result set；Obtain the complete of each multimedia resource in the first result set
Portion's information labels, according to all information label of each multimedia resource in filtering words and phrases and the first result set to the first result set
It is filtered, obtains the second result set.
In one possible implementation, this method further include:
Feature extraction is carried out to the content of multimedia resource, obtains key message；
Semantic analysis is carried out to extracted key message, is determined between key message and the text information of natural language
Incidence relation；
Each information labels and its corresponding degree of correlation index of multimedia resource are determined according to the incidence relation, wherein
The corresponding degree of correlation index of one information labels is used to indicate the degree of correlation of the information labels Yu its multimedia resource identified.
In one possible implementation, this method further include:
According to the information labels of each multimedia resource and its corresponding degree of correlation index, foundation obtains the pass of multimedia resource
Key information is associated with lexicon.
In one possible implementation, feature extraction is carried out to the content of multimedia resource, obtains key message, wrapped
Include following at least one mode:
Personage in multimedia resource is identified, personage's key message is obtained；
Audio in multimedia resource is identified, audio key message is obtained；
Scene in multimedia resource is identified, scene key message is obtained；
Place in multimedia resource is identified, place key message is obtained；
Article in multimedia resource is identified, article key message is obtained.
According to another aspect of the present disclosure, a kind of multimedia resource searcher is provided, comprising:
First semantic module obtains each pass relevant to search words and phrases for carrying out semantic analysis to search words and phrases
Join words and phrases；
Information labels searching module, for searching the information labels of each multimedia resource related with each association sentence, institute
The information labels for stating multimedia resource include the text information of natural language associated with the key message of multimedia resource, institute
The key message for stating multimedia resource is to carry out feature extraction to the content of multimedia resource to obtain；
Search result determining module determines included by search result for the information labels according to each multimedia resource
Multimedia resource.
In one possible implementation, first semantic module is also used to carry out semantic point to search words and phrases
Analysis obtains and the search relevant each retrieval words and phrases of words and phrases and each filtering words and phrases.
In one possible implementation, the information labels searching module is also used to the key letter in multimedia resource
In breath association lexicon, the first information label of each multimedia resource related with each retrieval words and phrases is searched；
Described search result determining module is also used to multimedia resource corresponding to first information label, as the first knot
Fruit collection；The all information label for obtaining each multimedia resource in the first result set, according in filtering words and phrases and the first result set
The all information label of each multimedia resource the first result set is filtered, obtain the second result set.
In one possible implementation, the device further include:
Characteristic extracting module carries out feature extraction for the content to multimedia resource, obtains key message；
Second semantic module determines key message and oneself for carrying out semantic analysis to extracted key message
Incidence relation between the text information of right language；
Information labels determining module, for determining each information labels of multimedia resource and its right according to the incidence relation
The degree of correlation index answered, wherein the corresponding degree of correlation index of an information labels is for indicating that the information labels are identified with it
Multimedia resource the degree of correlation.
In one possible implementation, the device further include:
Association lexicon establishes module, for being referred to according to the information labels and its corresponding degree of correlation of each multimedia resource
Number is established and obtains the key message association lexicon of multimedia resource.
In one possible implementation, the characteristic extracting module includes at least one of following submodule:
Person recognition submodule obtains personage's key message for identifying to the personage in multimedia resource；
Audio identification submodule obtains audio key message for identifying to the audio in multimedia resource；
Scene Recognition submodule obtains scene key message for identifying to the scene in multimedia resource；
Place identifies that submodule obtains place key message for identifying to the place in multimedia resource；
Article identifies that submodule obtains article key message for identifying to the article in multimedia resource.
According to another aspect of the present disclosure, a kind of multimedia resource searcher is provided, comprising: processor；For depositing
Store up the memory of processor-executable instruction；Wherein, the processor is configured to executing the above method.
According to another aspect of the present disclosure, a kind of non-volatile computer readable storage medium storing program for executing is provided, is stored thereon with
Computer program instructions, wherein the computer program instructions realize the above method when being executed by processor.
In the disclosure, the information labels of each multimedia resource include it is associated with the key message of multimedia resource from
The text information of right language, the key message of multimedia resource are to carry out feature extraction to the content of multimedia resource to obtain,
The associate text information of the content of multimedia resource and Human Natural Language can be got up.Therefore, in search, to search term
Sentence carries out semantic analysis and obtains association sentence, can search the information labels of corresponding multimedia resource, and then more accurately,
It is quickly obtained the content of required multimedia resource.
According to below with reference to the accompanying drawings to detailed description of illustrative embodiments, the other feature and aspect of the disclosure will become
It is clear.
Detailed description of the invention
Comprising in the description and constituting the attached drawing of part of specification and specification together illustrates the disclosure
Exemplary embodiment, feature and aspect, and for explaining the principles of this disclosure.
Fig. 1 shows the flow chart of the multimedia resource searching method according to one embodiment of the disclosure.
Fig. 2 shows the flow charts according to the multimedia resource searching method of another embodiment of the disclosure.
Fig. 3 shows the schematic diagram of the application scenarios of the multimedia resource searching method according to another embodiment of the disclosure.
Fig. 4 shows the block diagram of the multimedia resource searcher according to one embodiment of the disclosure.
Fig. 5 shows the block diagram of the multimedia resource searcher according to another embodiment of the disclosure.
Fig. 6 shows the block diagram of the multimedia resource searcher according to another embodiment of the disclosure.
Specific embodiment
Various exemplary embodiments, feature and the aspect of the disclosure are described in detail below with reference to attached drawing.It is identical in attached drawing
Appended drawing reference indicate element functionally identical or similar.Although the various aspects of embodiment are shown in the attached drawings, remove
It non-specifically points out, it is not necessary to attached drawing drawn to scale.
Dedicated word " exemplary " means " being used as example, embodiment or illustrative " herein.Here as " exemplary "
Illustrated any embodiment should not necessarily be construed as preferred or advantageous over other embodiments.
In addition, giving numerous details in specific embodiment below to better illustrate the disclosure.
It will be appreciated by those skilled in the art that without certain details, the disclosure equally be can be implemented.In some instances, for
Method, means, element and circuit well known to those skilled in the art are not described in detail, in order to highlight the purport of the disclosure.
Fig. 1 shows the flow chart of the multimedia resource searching method according to one embodiment of the disclosure.As shown in Figure 1, this is more
Media resource searching method includes:
Step 101 carries out semantic analysis to search words and phrases, obtains each association words and phrases relevant to search words and phrases.
Step 102, the information labels for searching each multimedia resource related with each association sentence, the multimedia resource
Information labels include the text information of natural language associated with the key message of multimedia resource, the multimedia resource
Key message is to carry out feature extraction to the content of multimedia resource to obtain.
Step 103, according to the information labels of each multimedia resource, determine multimedia resource included by search result.
In the disclosure, multimedia resource include but is not limited to one of resources such as video, picture, audio, text or
A variety of combinations.By taking video as an example, the content of video includes picture, sound, picture etc. in video.To in multimedia resource
Hold and carry out feature extraction, the key message such as personage, audio, scene, place can be obtained.Language is carried out to these key messages
Justice analysis, can obtain the text information of relevant natural language.Then, each multimedia is set further according to these text informations
The information labels of resource, in order to later use, these information labels are scanned for.Wherein, according to the letter of multiple multimedia resources
Label is ceased, the key message that can establish in multimedia resource is associated with lexicon.As shown in figure 3, the crucial letter of multimedia resource
Breath association lexicon can be stored in video features library.
In one possible implementation, step 101 includes: to carry out semantic analysis to search words and phrases, obtains and searches for
The relevant each retrieval words and phrases of words and phrases and each filtering words and phrases.
As shown in figure 3, search when, using include the mankind natural language between incidence relation semantic knowledge-base,
The search words and phrases for inputting or selecting to user carry out semantic analysis, obtain each association words and phrases relevant to the search words and phrases.These
Association words and phrases include but is not limited to the different clause expression etc. of synonym, antonym, near synonym, dialect, identical semanteme.Also,
It can also classify to these association words and phrases.Some words and phrases for retrieval referred to as retrieval words and phrases, such as synonym, nearly justice
Word, dialect etc..Some words and phrases for filtering referred to as filtering words and phrases, such as antonym etc..
In one possible implementation, step 102 includes: the key message association lexicon in multimedia resource
In, search the first information label of each multimedia resource related with each retrieval words and phrases.
Step 103 includes: by multimedia resource corresponding to first information label, as the first result set；Obtain first
The all information label of each multimedia resource in result set, according to each multimedia resource in filtering words and phrases and the first result set
All information label the first result set is filtered, obtain the second result set.
For example, obtaining the conjunctive words such as " happiness ", " being so happy as to weep ", " sad " according to search term " happy ".Wherein, it " opens
The heart ", " happiness ", " being so happy as to weep " belong to term, and " sad " belongs to filter word.In key message association lexicon, according to
Information labels " happy ", " happiness ", " being so happy as to weep ", search video A, B, C.Wherein, all information label of video A includes
The all information label of " happy ", " happiness ", " joyful ", video B includes " happy ", " being so happy as to weep ", " sad ", video C's
All information label includes " sad ", " pain ".It at this moment, can be using video A, B, C as the first result set.It is then possible to root
Video B, C are filtered out from the first result set according to " sad ", the second obtained result set includes video A.
In addition, each label information of each multimedia resource can also have correspondence in key message association lexicon
Degree of correlation index.In a step 102, each first information label that can also find each each multimedia resource is corresponding
Degree of correlation index.In step 103, label and its corresponding degree of correlation index the first result set packet can be determined according to the first information
The multimedia resource included.It then, can be according to the complete of each multimedia resource in filtering words and phrases, the first result set in filtering
Portion's information labels and its corresponding degree of correlation index, are filtered the first result set, obtain the second result set.Wherein, it retrieves
The degree of correlation index of words and phrases and multimedia resource is properly termed as matching degree index, and the degree of correlation for filtering vocabulary and multimedia resource refers to
Number is properly termed as filtering degree index,
Example is connected, the first result set includes video A, B, C.In all information label of video B, the degree of correlation of " happy " refers to
Number is 0.5, and the degree of correlation index of " being so happy as to weep " is 0.8, and the degree of correlation index of " sad " is 0.1.The all information mark of video C
In label, the degree of correlation index of " sad " is 0.8, and the degree of correlation index of " pain " is 0.7.In filtering, although finding video B
With include information labels " sad " in C, but since the degree of correlation index of video B is smaller, can not have to filter out video B.
The second result set obtained in this way includes video A and B.
In the present embodiment, the information labels of each multimedia resource include associated with the key message of multimedia resource
The text information of natural language, the key message of multimedia resource are to carry out feature extraction to the content of multimedia resource to obtain
, the associate text information of the content of multimedia resource and Human Natural Language can be got up.Therefore, in search, to searching
Rope words and phrases carry out semantic analysis and obtain association sentence, can search the information labels of corresponding multimedia resource, and then obtain
The content of required multimedia resource.
Fig. 2 shows the flow charts according to the multimedia resource searching method of another embodiment of the disclosure.As shown in Fig. 2, with
Above-described embodiment the difference is that, the multimedia resource searching method further include:
Step 201 carries out feature extraction to the content of multimedia resource, obtains key message；
Step 202 carries out semantic analysis to extracted key message, determines the text envelope of key message and natural language
Incidence relation between breath；
Step 203 determines that each information labels of multimedia resource and its corresponding degree of correlation refer to according to the incidence relation
Number, wherein the corresponding degree of correlation index of an information labels is used to indicate the information labels and its multimedia resource identified
The degree of correlation.
In the disclosure, feature extraction is carried out to the content of multimedia resource, can obtain for example personage, audio, scene,
The key messages such as place.Semantic analysis is carried out to these key messages using semantic knowledge-base, relevant natural language can be obtained
The text information of speech.
In one possible implementation, step 201 comprises at least one of the following mode:
Personage in multimedia resource is identified, personage's key message is obtained；
Audio in multimedia resource is identified, audio key message is obtained；
Scene in multimedia resource is identified, scene key message is obtained；
Place in multimedia resource is identified, place key message is obtained；
Article in multimedia resource is identified, article key message is obtained.
For example, identifying that the face feature of personage belongs to " smiling face " this key for being embodied expression from the picture of video
Information.Further, " smiling face " can be associated with to the abstract expression of the moods such as " happy ", " happiness " in semantic knowledge-base
Text information.
For another example, march, symphony etc. are identified from audio, it can also be according to march, friendship in semantic knowledge-base
Ring the text information that happy feature is associated with " magnificence ", " feeling is aroused " these emotion expression services.
Then, the information labels of each multimedia resource are set further according to these text informations, in order to later use this
A little information labels scan for.
In one possible implementation, this method further include:
According to the information labels of each multimedia resource and its corresponding degree of correlation index, foundation obtains the pass of multimedia resource
Key information is associated with lexicon.
As shown in figure 3, may include the mark (ID) of video in video material library, title (title), director, performer, interior
Hold the essential informations such as introduction.E.g., including film, TV play the introduction general content that includes.
It may include the tool such as the color extracted from video, article, sound, song as intuitive in video material tag library
Part key message.These key messages belong in Human Natural Language is embodied expression without mood.
To some video in the essential information in video material library and the key message of video material tag library, known by semanteme
Know a degree of semantic analysis of library progress and various intuitively to abstract association.In this way, available be associated with the video
Various information labels and each information labels and the video degree of correlation index etc..By the video and information labels, related
The incidence relations such as degree index are saved in key message association lexicon.This key message association lexicon can be set in video
In feature database.
It, then, can should in video features library for example, have in one section of video red knife (video material tag library)
Plus such as " homicide ", " lethal weapon ", these pass through thinking judgement (such as semantic knowledge-base is analyzed to obtain) in the information labels of video
Corresponding relationship afterwards.The label information of this section of video may include " homicide ", feature as " lethal weapon ".In addition, if video
In material database, this video further includes " Li Lianjie " this performer, then what can be come out with comprehensive descision is " acrobatic fighting film ".If depending on
The video is also associated with " war drum " in frequency material tag library, this section of video may not be swordsman, but ancient war piece.
For another example, the key message extracted from one section of video includes: " smiling face ", " sky ", " thunder ".These are crucial
Information can be saved in video material tag library." happy " is obtained, according to " day according to " smiling face " association in semantic knowledge-base
It is empty " it is associated with to obtain " rainy day " with " thunder ".Further, it can also be associated with according to " happy " with " rainy day " in semantic knowledge-base
Obtain word, phrase, short sentence as similar " having a long-felt need satisfied ".In this way, saving the video in video features library and " opening
The heart ", " rainy day ", " having a long-felt need satisfied " incidence relation.
In video features library, the associated information labels (being referred to as characteristic key words) of each video energy are The more the better.
Each information labels have a degree of correlation index (or being confidence score, scoring is higher to illustrate that identification is more accurate).
In the present embodiment, feature extraction is carried out to the content of multimedia resource and obtains key message, to extracted pass
Key information carries out semantic analysis, determines the incidence relation between key message and the text information of natural language, can be by more matchmakers
The content of body resource and the associate text information of Human Natural Language get up.Therefore, in search, search words and phrases are carried out semantic
Analysis obtains association sentence, can search the information labels of corresponding multimedia resource, and then obtains required multimedia money
The content in source.
Using the multimedia resource searching method of the disclosure, it can more accurately meet creator and specific multimedia is provided
The search need of the content in source.Regardless of search words and phrases belong to be embodied expression still fall within abstract expression, pass through semantic analysis
After being associated with, it can search for obtain the content for the multimedia resource being consistent with search words and phrases.
Using example: as shown in figure 3, by taking video as an example, the multimedia resource searching method the following steps are included:
Step 1 extracts key message from video.The key message of extraction can be stored in video material tag library.
Recognition of face, voice recognition, scene Recognition, place identification, article identification such as are carried out to video, example is as follows:
A) personage's static state identifies: the key features such as identifying characters name, age from video, dresss up.Such as identify people
Whether the dress ornament of object has beard, wrinkle, mole, double-edged eyelid etc., if there is beard can also identify beard type etc..
B) movement, expression of personage etc. personage's Dynamic Recognition: are identified from video.
C) audio identification: identify that (such as news joins with happy lines for music, song, well-known program from video
Broadcast opening remarks etc.), natural sound such as thunder and lightning, voice (such as pitch, thickness, emergency audio feature information), lines etc..
D) scene Recognition: identifying light (such as color, light intensity) from video, building attribute (such as high building, audience hall,
Square, market, cavern etc.).
E) place identifies: pinpoint terrestrial reference geography geologic feature or can build on map from identifying in video
Build object etc..
F) article identifies: the common items such as daily necessity, electric appliance, the various vehicles, weapon are identified from video, with
And specific certain articles, such as the exclusive stage property in films and television programs.
Step 2 establishes semantic knowledge map.Understood according to the image of Human Natural Language, the relevance between words and phrases etc.,
Establish key message association lexicon, as semantic knowledge map.Using semantic knowledge-base to the video in video material library
Key message in essential information and video material tag library carries out semantic analysis and association, obtains the information labels and phase of video
Pass degree index, and by the information labels of video and degree of correlation index mapping into semantic knowledge map, it can be for subsequent quick inspection
Rope uses.Semantic knowledge map can be saved in video features library.
Step 3, intelligent search.
As shown in figure 3, carrying out semantic analysis to search term, sentence, characteristic information is extracted.Then, from the pass in video features library
The characteristic value namely information labels inquired all relative words in key information association lexicon, describe clause.Further according to these spies
Value indicative is retrieved from video features library and obtains required result set.Then result set is filtered.Finally according to matching degree into
Row sequence, more accurately result set is presented to searchers for output.
Step 3a, association map, including but not limited to synonym, antonym, near synonym, side are constructed to Human Natural Language
Speech, identical semanteme different clause express etc., some cyberspeaks etc. can also translate into natural language, such as by search term
" eating soil " translates into " out of funds ".As shown in figure 3, natural language association map can be saved in semantic database.
Step 3b, according to search term, sentence from natural language association map in inquire all qualified association vocabulary,
Clause (hereinafter referred to as " words and phrases " or " association words and phrases ").For example, being determined using semantic knowledge-base and searching for the related pass of words and phrases
Join words and phrases.
Association words and phrases are divided into two parts, for retrieving (hereinafter referred to as " retrieval words and phrases "), another part is used for a part
Filtering (hereinafter referred to as " filtering words and phrases ").
Step 3c, according to retrieval words and phrases, PRELIMINARY RESULTS collection is retrieved from video features library.PRELIMINARY RESULTS collection includes a system
The corresponding matching degree index of the information labels of column video and each video.
Step 3d, it is filtered according to filter word sentence pair PRELIMINARY RESULTS collection.The principle of filtering is that the same video includes
In all information labels, the part identical with retrieval words and phrases might have, while being also possible to containing other information labels and mistake
It is related to filter words and phrases.According to the available filtering degree index of degree of agreement of label and filtering words and phrases, when filtering degree index is greater than one
When determining threshold value, then the video should be filtered.Unsorted result set is obtained after filtering.Unsorted result set includes a series of views
Frequency material, matching degree index and filtering degree index.
Step 3e, it is ranked up according to the video in matching degree index and filtering degree exponent pair result set, is most terminated
Fruit collection.
In addition, can also be done using advanced way of search to search target and further be definitely defined in search process.
Such as: search words and phrases are " lonely cat-Huang cat ", can filter out the cat of yellow from the result set of " lonely cat ".For another example,
It is " strong middle-aged man-title: hero " with search words and phrases, can be filtered out from the result set of " strong middle-aged man "
It include the result of " hero " in title.
The disclosure is different from traditional search engines to the processing mode of audio video searching, can be to the language that audio-video is included
The expression such as justice, perception, image are matched, and are searched out from mass data and are met expected result.
Using the multimedia resource searching method of the disclosure, suitable element can be found in the materials such as video, audio
Material.The disclosure is some abstract tables relevant to emotion expression service, psychological feelings, sensory experience especially suitable for search target
It reaches, such as search term is " melancholy ", " dim ", " ear-piercing ", the description of " terrified " adjective, or " falling suddenly " etc.
Adverbial description.
Using the scheme of the disclosure, it can preferably meet creator in various search needs, such as:
1) what creator faced is massive video material, and wherein most is not watch, the process of creation
In need therefrom to find out the video clip or picture for meeting oneself demand again.
2) duration of audio visual work is relatively long, and the possibility that creator needs is several seconds included in audio-video
A camera lens, some picture of clock.
3) theme, description text can summarize the entirety of video content, mainly to describe based on story train of thought.But
For be directed to scene, camera lens, lack description and index information.For example it is showed in a comedy movie TV play sad
Etc. negative emotions picture, may more meet in expression way and effect creator intention.
4) the relevant number of person of movie and television play is huge, and performer's name and role name etc. mix, and have some names in official
It is unwritten in Fang Wenben.For example Zhou Xing early stage of speeding plays a bit role the camera lens etc. of appearance in Hero Shooting Vulture.And these are believed
Breath can be extracted by audio-video identification technology, be used for search.
5) in video figure image, place, building humanized abstractdesription, such as " keep the hard of dense whiskers
Bright middle-aged male is sitting on the ground extremely grievedly ", " glazed tiles on audience hall top reflect magnificent gold under irradiation by sunlight
Light ", " original sunny day overcasts suddenly in the air, lightning accompanied by peals of thunder " etc., usually will not in the verbal description of video
Occur, relevant text will not do complete description to details in the process in webpage.
6) audio-frequency information can not be abstracted into verbal description, such as " other movie and television plays that the voice-over actor of Sun Wukong matched ",
Or " raising a cry of warning " it is corresponding may be some relatively quiet picture offscreen voice, more more complicated is " outside window small
Girl is raised a cry of warning while running: ' kindling！' " this other than lines, further comprise the identification to various sound.
Fig. 4 shows the block diagram of the multimedia resource searcher according to one embodiment of the disclosure.As shown in figure 4, more matchmakers
Body resource searching device includes:
First semantic module 41 obtains relevant to search words and phrases each for carrying out semantic analysis to search words and phrases
It is associated with words and phrases；
Information labels searching module 43, for searching the information labels of each multimedia resource related with each association sentence,
The information labels of the multimedia resource include the text information of natural language associated with the key message of multimedia resource,
The key message of the multimedia resource is to carry out feature extraction to the content of multimedia resource to obtain；
Search result determining module 45 determines included by search result for the information labels according to each multimedia resource
Multimedia resource.
Fig. 5 shows the block diagram of the multimedia resource searcher according to another embodiment of the disclosure.As shown in figure 5, with upper
One embodiment the difference is that, the first semantic module 41 of the multimedia resource searcher is also used to search term
Sentence carries out semantic analysis, obtains and the search relevant each retrieval words and phrases of words and phrases and each filtering words and phrases.
In one possible implementation, the information labels searching module 43 is also used to the key in multimedia resource
In information association lexicon, the first information label of each multimedia resource related with each retrieval words and phrases is searched；
Described search result determining module 45 is also used to by multimedia resource corresponding to first information label, as first
Result set；The all information label for obtaining each multimedia resource in the first result set, according to filtering words and phrases and the first result set
In all information label of each multimedia resource the first result set is filtered, obtain the second result set.
In one possible implementation, the device further include:
Characteristic extracting module 51 carries out feature extraction for the content to multimedia resource, obtains key message；
Second semantic module 53, for extracted key message carry out semantic analysis, determine key message with
Incidence relation between the text information of natural language；
Information labels determining module 55, for determined according to the incidence relation multimedia resource each information labels and its
Corresponding degree of correlation index, wherein the corresponding degree of correlation index of an information labels is for indicating that the information labels are marked with it
The degree of correlation of the multimedia resource of knowledge.
In one possible implementation, the device further include:
Association lexicon establishes module 57, for being referred to according to the information labels and its corresponding degree of correlation of each multimedia resource
Number is established and obtains the key message association lexicon of multimedia resource.
In one possible implementation, the characteristic extracting module 51 includes at least one of following submodule:
Person recognition submodule obtains personage's key message for identifying to the personage in multimedia resource；
Audio identification submodule obtains audio key message for identifying to the audio in multimedia resource；
Scene Recognition submodule obtains scene key message for identifying to the scene in multimedia resource；
Place identifies that submodule obtains place key message for identifying to the place in multimedia resource；
Article identifies that submodule obtains article key message for identifying to the article in multimedia resource.
About the device in above-described embodiment, wherein modules execute the concrete mode of operation in related this method
Embodiment in be described in detail, no detailed explanation will be given here.
Fig. 6 shows the block diagram of the multimedia resource searcher according to another embodiment of the disclosure.For example, device 1900 can
To be provided as a server.Referring to Fig. 6, it further comprises one or more places that device 1900, which includes processing component 1922,
Manage device and memory resource represented by a memory 1932, for store can by the instruction of the execution of processing component 1922,
Such as application program.The application program stored in memory 1932 may include it is one or more each correspond to one
The module of group instruction.In addition, processing component 1922 is configured as executing instruction, to execute the above method.
Device 1900 can also include that a power supply module 1926 be configured as the power management of executive device 1900, and one
Wired or wireless network interface 1950 is configured as device 1900 being connected to network and input and output (I/O) interface
1958.Device 1900 can be operated based on the operating system for being stored in memory 1932, such as Windows ServerTM, Mac
OS XTM, UnixTM, LinuxTM, FreeBSDTM or similar.
In the exemplary embodiment, a kind of non-volatile computer readable storage medium storing program for executing is additionally provided, for example including calculating
The memory 1932 of machine program instruction, above-mentioned computer program instructions can be executed by the processing component 1922 of device 1900 to complete
The above method.
The disclosure can be system, method and/or computer program product.Computer program product may include computer
Readable storage medium storing program for executing, containing for making processor realize the computer-readable program instructions of various aspects of the disclosure.
Computer readable storage medium, which can be, can keep and store the tangible of the instruction used by instruction execution equipment
Equipment.Computer readable storage medium for example can be-- but it is not limited to-- storage device electric, magnetic storage apparatus, optical storage
Equipment, electric magnetic storage apparatus, semiconductor memory apparatus or above-mentioned any appropriate combination.Computer readable storage medium
More specific example (non exhaustive list) includes: portable computer diskette, hard disk, random access memory (RAM), read-only deposits
It is reservoir (ROM), erasable programmable read only memory (EPROM or flash memory), static random access memory (SRAM), portable
Compact disk read-only memory (CD-ROM), digital versatile disc (DVD), memory stick, floppy disk, mechanical coding equipment, for example thereon
It is stored with punch card or groove internal projection structure and the above-mentioned any appropriate combination of instruction.Calculating used herein above
Machine readable storage medium storing program for executing is not interpreted that instantaneous signal itself, the electromagnetic wave of such as radio wave or other Free propagations lead to
It crosses the electromagnetic wave (for example, the light pulse for passing through fiber optic cables) of waveguide or the propagation of other transmission mediums or is transmitted by electric wire
Electric signal.
Computer-readable program instructions as described herein can be downloaded to from computer readable storage medium it is each calculate/
Processing equipment, or outer computer or outer is downloaded to by network, such as internet, local area network, wide area network and/or wireless network
Portion stores equipment.Network may include copper transmission cable, optical fiber transmission, wireless transmission, router, firewall, interchanger, gateway
Computer and/or Edge Server.Adapter or network interface in each calculating/processing equipment are received from network to be counted
Calculation machine readable program instructions, and the computer-readable program instructions are forwarded, for the meter being stored in each calculating/processing equipment
In calculation machine readable storage medium storing program for executing.
Computer program instructions for executing disclosure operation can be assembly instruction, instruction set architecture (ISA) instructs,
Machine instruction, machine-dependent instructions, microcode, firmware instructions, condition setup data or with one or more programming languages
The source code or object code that any combination is write, the programming language include the programming language-of object-oriented such as
Smalltalk, C++ etc., and conventional procedural programming languages-such as " C " language or similar programming language.Computer
Readable program instructions can be executed fully on the user computer, partly execute on the user computer, be only as one
Vertical software package executes, part executes on the remote computer or completely in remote computer on the user computer for part
Or it is executed on server.In situations involving remote computers, remote computer can pass through network-packet of any kind
It includes local area network (LAN) or wide area network (WAN)-is connected to subscriber computer, or, it may be connected to outer computer (such as benefit
It is connected with ISP by internet).In some embodiments, by utilizing computer-readable program instructions
Status information carry out personalized customization electronic circuit, such as programmable logic circuit, field programmable gate array (FPGA) or can
Programmed logic array (PLA) (PLA), the electronic circuit can execute computer-readable program instructions, to realize each side of the disclosure
Face.
Referring herein to according to the flow chart of the method, apparatus (system) of the embodiment of the present disclosure and computer program product and/
Or block diagram describes various aspects of the disclosure.It should be appreciated that flowchart and or block diagram each box and flow chart and/
Or in block diagram each box combination, can be realized by computer-readable program instructions.
These computer-readable program instructions can be supplied to general purpose computer, special purpose computer or other programmable datas
The processor of processing unit, so that a kind of machine is produced, so that these instructions are passing through computer or other programmable datas
When the processor of processing unit executes, function specified in one or more boxes in implementation flow chart and/or block diagram is produced
The device of energy/movement.These computer-readable program instructions can also be stored in a computer-readable storage medium, these refer to
It enables so that computer, programmable data processing unit and/or other equipment work in a specific way, thus, it is stored with instruction
Computer-readable medium then includes a manufacture comprising in one or more boxes in implementation flow chart and/or block diagram
The instruction of the various aspects of defined function action.
Computer-readable program instructions can also be loaded into computer, other programmable data processing units or other
In equipment, so that series of operation steps are executed in computer, other programmable data processing units or other equipment, to produce
Raw computer implemented process, so that executed in computer, other programmable data processing units or other equipment
Instruct function action specified in one or more boxes in implementation flow chart and/or block diagram.
The flow chart and block diagram in the drawings show system, method and the computer journeys according to multiple embodiments of the disclosure
The architecture, function and operation in the cards of sequence product.In this regard, each box in flowchart or block diagram can generation
One module of table, program segment or a part of instruction, the module, program segment or a part of instruction include one or more use
The executable instruction of the logic function as defined in realizing.In some implementations as replacements, function marked in the box
It can occur in a different order than that indicated in the drawings.For example, two continuous boxes can actually be held substantially in parallel
Row, they can also be executed in the opposite order sometimes, and this depends on the function involved.It is also noted that block diagram and/or
The combination of each box in flow chart and the box in block diagram and or flow chart, can the function as defined in executing or dynamic
The dedicated hardware based system made is realized, or can be realized using a combination of dedicated hardware and computer instructions.
The presently disclosed embodiments is described above, above description is exemplary, and non-exclusive, and
It is not limited to disclosed each embodiment.Without departing from the scope and spirit of illustrated each embodiment, for this skill
Many modifications and changes are obvious for the those of ordinary skill in art field.The selection of term used herein, purport
In the principle, practical application or technological improvement to the technology in market for best explaining each embodiment, or lead this technology
Other those of ordinary skill in domain can understand each embodiment disclosed herein.
Claims (14)
1. a kind of multimedia resource searching method characterized by comprising
Semantic analysis is carried out to search words and phrases, obtains each association words and phrases relevant to search words and phrases；
The information labels of each multimedia resource related with each association sentence are searched, the information labels of the multimedia resource include
The text information of natural language associated with the key message of multimedia resource, the key message of the multimedia resource are pair
The content of multimedia resource carries out what feature extraction obtained；
According to the information labels of each multimedia resource, multimedia resource included by search result is determined.
2. being obtained and search term the method according to claim 1, wherein carrying out semantic analysis to search words and phrases
The relevant each association words and phrases of sentence, comprising:
Semantic analysis is carried out to search words and phrases, is obtained and the search relevant each retrieval words and phrases of words and phrases and each filtering words and phrases.
3. according to the method described in claim 2, it is characterized in that, searching each multimedia resource related with each association sentence
Information labels, comprising: in the key message association lexicon of multimedia resource, search each more matchmakers related with each retrieval words and phrases
The first information label of body resource；
According to the information labels of each multimedia resource, multimedia resource included by search result is determined, comprising: by the first information
Multimedia resource corresponding to label, as the first result set；Obtain whole letters of each multimedia resource in the first result set
Label is ceased, the first result set is carried out according to all information label of each multimedia resource in filtering words and phrases and the first result set
Filtering, obtains the second result set.
4. the method according to claim 1, wherein further include:
Feature extraction is carried out to the content of multimedia resource, obtains key message；
Semantic analysis is carried out to extracted key message, determines being associated between key message and the text information of natural language
Relationship；
Each information labels and its corresponding degree of correlation index of multimedia resource are determined according to the incidence relation, wherein one
The corresponding degree of correlation index of information labels is used to indicate the degree of correlation of the information labels Yu its multimedia resource identified.
5. according to the method described in claim 4, it is characterized by further comprising:
According to the information labels of each multimedia resource and its corresponding degree of correlation index, establishes and obtain the crucial letter of multimedia resource
Breath association lexicon.
6. according to the method described in claim 4, obtaining it is characterized in that, carry out feature extraction to the content of multimedia resource
Key message comprises at least one of the following mode:
Personage in multimedia resource is identified, personage's key message is obtained；
Audio in multimedia resource is identified, audio key message is obtained；
Scene in multimedia resource is identified, scene key message is obtained；
Place in multimedia resource is identified, place key message is obtained；
Article in multimedia resource is identified, article key message is obtained.
7. a kind of multimedia resource searcher characterized by comprising
First semantic module obtains each conjunctive word relevant to search words and phrases for carrying out semantic analysis to search words and phrases
Sentence；
Information labels searching module is described more for searching the information labels of each multimedia resource related with each association sentence
The information labels of media resource include the text information of natural language associated with the key message of multimedia resource, described more
The key message of media resource is to carry out feature extraction to the content of multimedia resource to obtain；
Search result determining module determines more matchmakers included by search result for the information labels according to each multimedia resource
Body resource.
8. device according to claim 7, which is characterized in that first semantic module is also used to search words and phrases
Semantic analysis is carried out, is obtained and the search relevant each retrieval words and phrases of words and phrases and each filtering words and phrases.
9. device according to claim 8, which is characterized in that the information labels searching module is also used to provide in multimedia
In the key message association lexicon in source, the first information label of each multimedia resource related with each retrieval words and phrases is searched；
Described search result determining module is also used to multimedia resource corresponding to first information label, as the first result
Collection；The all information label for obtaining each multimedia resource in the first result set, according in filtering words and phrases and the first result set
The all information label of each multimedia resource is filtered the first result set, obtains the second result set.
10. device according to claim 7, which is characterized in that further include:
Characteristic extracting module carries out feature extraction for the content to multimedia resource, obtains key message；
Second semantic module determines key message and natural language for carrying out semantic analysis to extracted key message
Incidence relation between the text information of speech；
Information labels determining module, for determining each information labels of multimedia resource and its corresponding according to the incidence relation
Degree of correlation index, wherein the corresponding degree of correlation index of an information labels is for indicating that it is more that the information labels and its are identified
The degree of correlation of media resource.
11. device according to claim 10, which is characterized in that further include:
Association lexicon establishes module, for the information labels and its corresponding degree of correlation index according to each multimedia resource, builds
The vertical key message association lexicon for obtaining multimedia resource.
12. device according to claim 10, which is characterized in that the characteristic extracting module include following submodule extremely
Few one kind:
Person recognition submodule obtains personage's key message for identifying to the personage in multimedia resource；
Audio identification submodule obtains audio key message for identifying to the audio in multimedia resource；
Scene Recognition submodule obtains scene key message for identifying to the scene in multimedia resource；
Place identifies that submodule obtains place key message for identifying to the place in multimedia resource；
Article identifies that submodule obtains article key message for identifying to the article in multimedia resource.
13. a kind of multimedia resource searcher characterized by comprising
Processor；
Memory for storage processor executable instruction；
Wherein, the processor is configured to: perform claim require any one of 1 to 6 described in method.
14. a kind of non-volatile computer readable storage medium storing program for executing, is stored thereon with computer program instructions, which is characterized in that institute
It states and realizes method described in any one of claim 1 to 6 when computer program instructions are executed by processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711444505.7A CN110019852A (en) | 2017-12-27 | 2017-12-27 | Multimedia resource searching method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711444505.7A CN110019852A (en) | 2017-12-27 | 2017-12-27 | Multimedia resource searching method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110019852A true CN110019852A (en) | 2019-07-16 |
Family
ID=67187073
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711444505.7A Pending CN110019852A (en) | 2017-12-27 | 2017-12-27 | Multimedia resource searching method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110019852A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110430476A (en) * | 2019-08-05 | 2019-11-08 | 广州华多网络科技有限公司 | Direct broadcasting room searching method, system, computer equipment and storage medium |
| CN110532404A (en) * | 2019-09-03 | 2019-12-03 | 北京百度网讯科技有限公司 | One provenance multimedia determines method, apparatus, equipment and storage medium |
| CN111159435A (en) * | 2019-12-27 | 2020-05-15 | 北大方正集团有限公司 | Multimedia resource processing method, system, terminal and computer readable storage medium |
| CN111639228A (en) * | 2020-05-29 | 2020-09-08 | 北京百度网讯科技有限公司 | Video retrieval method, device, equipment and storage medium |
| CN113469298A (en) * | 2021-09-03 | 2021-10-01 | 北京达佳互联信息技术有限公司 | Model training method and resource recommendation method |
| CN113641857A (en) * | 2021-08-13 | 2021-11-12 | 三星电子（中国）研发中心 | Visual media personalized search method and device |
| CN113934869A (en) * | 2021-09-23 | 2022-01-14 | 阿里云计算有限公司 | Database construction method, multimedia file retrieval method and device |
| CN114625897A (en) * | 2022-03-21 | 2022-06-14 | 腾讯科技（深圳）有限公司 | Multimedia resource processing method and device, electronic equipment and storage medium |
| CN116095422A (en) * | 2022-12-15 | 2023-05-09 | 上海幻电信息科技有限公司 | Video generation method and device, storage medium and electronic equipment |
| CN116992834A (en) * | 2023-09-27 | 2023-11-03 | 国高信息技术产业研究院(山东)集团有限公司 | Auxiliary writing method, system, medium and electronic equipment based on artificial intelligence |
| US12038968B2 (en) | 2021-08-13 | 2024-07-16 | Samsung Electronics Co., Ltd. | Method and device for personalized search of visual media |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101251844A (en) * | 2007-02-21 | 2008-08-27 | 富士胶片株式会社 | Apparatus and method for retrieval of contents |
| CN101901249A (en) * | 2009-05-26 | 2010-12-01 | 复旦大学 | Text-based query expansion and sort method in image retrieval |
| CN103106287A (en) * | 2013-03-06 | 2013-05-15 | 深圳市宜搜科技发展有限公司 | Processing method and processing system for retrieving sentences by user |
| CN103440284A (en) * | 2013-08-14 | 2013-12-11 | 郭克华 | Multimedia storage and search method supporting cross-type semantic search |
| CN104915433A (en) * | 2015-06-24 | 2015-09-16 | 宁波工程学院 | Method for searching for film and television video |
-
2017
- 2017-12-27 CN CN201711444505.7A patent/CN110019852A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101251844A (en) * | 2007-02-21 | 2008-08-27 | 富士胶片株式会社 | Apparatus and method for retrieval of contents |
| CN101901249A (en) * | 2009-05-26 | 2010-12-01 | 复旦大学 | Text-based query expansion and sort method in image retrieval |
| CN103106287A (en) * | 2013-03-06 | 2013-05-15 | 深圳市宜搜科技发展有限公司 | Processing method and processing system for retrieving sentences by user |
| CN103440284A (en) * | 2013-08-14 | 2013-12-11 | 郭克华 | Multimedia storage and search method supporting cross-type semantic search |
| CN104915433A (en) * | 2015-06-24 | 2015-09-16 | 宁波工程学院 | Method for searching for film and television video |
Non-Patent Citations (1)
| Title |
|---|
| 栾悉道 等: "多媒体语义模型研究进展", 《计算机科学》 * |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110430476A (en) * | 2019-08-05 | 2019-11-08 | 广州华多网络科技有限公司 | Direct broadcasting room searching method, system, computer equipment and storage medium |
| CN110430476B (en) * | 2019-08-05 | 2021-12-28 | 广州方硅信息技术有限公司 | Live broadcast room searching method, system, computer equipment and storage medium |
| CN110532404B (en) * | 2019-09-03 | 2023-08-04 | 北京百度网讯科技有限公司 | Source multimedia determining method, device, equipment and storage medium |
| CN110532404A (en) * | 2019-09-03 | 2019-12-03 | 北京百度网讯科技有限公司 | One provenance multimedia determines method, apparatus, equipment and storage medium |
| CN111159435A (en) * | 2019-12-27 | 2020-05-15 | 北大方正集团有限公司 | Multimedia resource processing method, system, terminal and computer readable storage medium |
| CN111159435B (en) * | 2019-12-27 | 2023-09-05 | 新方正控股发展有限责任公司 | Multimedia resource processing method, system, terminal and computer readable storage medium |
| CN111639228A (en) * | 2020-05-29 | 2020-09-08 | 北京百度网讯科技有限公司 | Video retrieval method, device, equipment and storage medium |
| CN113641857A (en) * | 2021-08-13 | 2021-11-12 | 三星电子（中国）研发中心 | Visual media personalized search method and device |
| US12038968B2 (en) | 2021-08-13 | 2024-07-16 | Samsung Electronics Co., Ltd. | Method and device for personalized search of visual media |
| CN113469298A (en) * | 2021-09-03 | 2021-10-01 | 北京达佳互联信息技术有限公司 | Model training method and resource recommendation method |
| CN113469298B (en) * | 2021-09-03 | 2022-01-21 | 北京达佳互联信息技术有限公司 | Model training method and resource recommendation method |
| CN113934869A (en) * | 2021-09-23 | 2022-01-14 | 阿里云计算有限公司 | Database construction method, multimedia file retrieval method and device |
| CN114625897A (en) * | 2022-03-21 | 2022-06-14 | 腾讯科技（深圳）有限公司 | Multimedia resource processing method and device, electronic equipment and storage medium |
| CN116095422A (en) * | 2022-12-15 | 2023-05-09 | 上海幻电信息科技有限公司 | Video generation method and device, storage medium and electronic equipment |
| CN116992834A (en) * | 2023-09-27 | 2023-11-03 | 国高信息技术产业研究院(山东)集团有限公司 | Auxiliary writing method, system, medium and electronic equipment based on artificial intelligence |
| CN116992834B (en) * | 2023-09-27 | 2024-01-05 | 国高信息技术产业研究院(山东)集团有限公司 | Auxiliary writing method, system, medium and electronic equipment based on artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110019852A (en) | Multimedia resource searching method and device | |
| US11475897B2 (en) | Method and apparatus for response using voice matching user category | |
| US10679063B2 (en) | Recognizing salient video events through learning-based multimodal analysis of visual features and audio-based analytics | |
| CN108604228B (en) | System and method for linguistic feature generation for multi-layered word representations | |
| CN103761261B (en) | A kind of media search method and device based on speech recognition | |
| CN103593363B (en) | The method for building up of video content index structure, video retrieval method and device | |
| CN110121116A (en) | Video generation method and device | |
| CN108833973A (en) | Extracting method, device and the computer equipment of video features | |
| US20140172419A1 (en) | System and method for generating personalized tag recommendations for tagging audio content | |
| US20110320197A1 (en) | Method for indexing multimedia information | |
| US11151191B2 (en) | Video content segmentation and search | |
| WO2007043679A1 (en) | Information processing device, and program | |
| CN109635295B (en) | Poetry retrieval method and system based on semantic analysis | |
| CN114254158B (en) | Video generation method and device, and neural network training method and device | |
| Vryzas et al. | Speech emotion recognition adapted to multimodal semantic repositories | |
| KR102619568B1 (en) | Voice based search for digital content in a network | |
| CN109271459A (en) | Chat robots and its implementation based on Lucene and grammer networks | |
| CN113392273A (en) | Video playing method and device, computer equipment and storage medium | |
| CN110121033A (en) | Video categorization and device | |
| CN112133406B (en) | Multi-mode emotion guidance method and system based on emotion maps and storage medium | |
| CN114996506B (en) | Corpus generation method, corpus generation device, electronic equipment and computer readable storage medium | |
| CN116092472A (en) | Speech synthesis method and synthesis system | |
| CN109376145B (en) | Method and device for establishing movie and television dialogue database and storage medium | |
| CN113689633B (en) | Scenic spot human-computer interaction method, device and system | |
| CN114547373A (en) | Method for intelligently identifying and searching programs based on audio |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20200507 Address after: 310052 room 508, floor 5, building 4, No. 699, Wangshang Road, Changhe street, Binjiang District, Hangzhou City, Zhejiang Province Applicant after: Alibaba (China) Co.,Ltd. Address before: 200241 room 1162, building 555, Dongchuan Road, Shanghai, Minhang District Applicant before: SHANGHAI QUAN TOODOU CULTURAL COMMUNICATION Co.,Ltd. |
|
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190716 |