Video source identification method and device

Abstract

The embodiment of the invention provides a video source identification method and device, computer equipment, a medium and a program product. Relates to the technical field of computers. The method comprises the following steps: acquiring a to-be-identified video; extracting a content text and multiple video frames from the video to be identified; performing text analysis on the content text through a large language model to obtain source information of the content text, the source information including a video name; performing face recognition on the plurality of video frames to obtain character information contained in the plurality of video frames; and determining a video source of the to-be-recognized video based on figure information obtained by face recognition and figure information contained in a target video corresponding to the video name. According to the technical scheme of the embodiment of the invention, whether the to-be-recognized video is the original video can be quickly and accurately recognized.

Description

Technical Field

[0001] Embodiments of the present application relate to the field of computer technologies, and in particular, to a method, device, computer device, computer-readable storage medium, and computer program product for identifying the source of a video. Background Art

[0002] In UGC (User Generated Content) platforms, it is often encountered that the videos uploaded by users involve copyright issues.

[0003] Traditional UGC platforms generally hire a large number of reviewers to review video content, including the review of copyright issues. However, since the content produced by users is very large, it is very resource-consuming to conduct copyright reviews manually, and the accuracy and real-time performance of manual reviews are also poor.

[0004] It should be noted that the above content is not necessarily prior art and is not used to limit the patent protection scope of the present application. Summary of the Invention

[0005] Embodiments of the present application provide a method, device, computer device, computer-readable storage medium, and computer program product for identifying the source of a video to solve or alleviate one or more of the above technical problems.

[0006] One aspect of embodiments of the present application provides a method for identifying the source of a video, the method including: Obtaining a video to be identified; Extracting content text and multiple video frames from the video to be identified; Performing text analysis on the content text through a large language model to obtain source information of the content text, the source information including a video name; Performing face recognition on the multiple video frames to obtain person information included in the multiple video frames; Determining the video source of the video to be identified based on the person information obtained by face recognition and the person information included in a target video corresponding to the video name.

[0007] Optionally, the extracting content text from the video to be identified includes: Converting the audio content in the video to be identified into content text by using automatic speech recognition technology.

[0008] Optionally, the method further includes: Detecting whether there is subtitle text in the video to be identified; In the case that the video to be recognized has subtitle text, extract the subtitle text from the video to be recognized, and use the extracted subtitle text as the content text.

[0009] Optionally, the performing face recognition on the multiple video frames to obtain the person information included in the multiple video frames includes: Perform face recognition on the multiple video frames respectively to obtain the person information included in each video frame; Perform statistical processing on the person information included in all video frames to obtain the quantities of multiple different person information; Delete the person information with a quantity lower than the preset quantity, and use the remaining person information as the person information included in the multiple video frames.

[0010] Optionally, the determining the video source of the video to be recognized based on the person information recognized by face recognition and the person information included in the target video corresponding to the video name includes: Determine the video source of the video to be recognized through the large language model based on the person information recognized by face recognition and the person information included in the target video corresponding to the video name.

[0011] Optionally, the source information includes multiple video names, and the method further includes: Combine the person information with the target videos corresponding to the multiple video names respectively to obtain the person information of multiple groups of target videos; The determining the video source of the video to be recognized through the large language model based on the person information recognized by face recognition and the person information included in the target video corresponding to the video name includes: Determine the video source of the video to be recognized through the large language model based on the matching degrees between the person information of the multiple groups of target videos and the person information included in the multiple target videos.

[0012] Optionally, the determining the video source of the video to be recognized through the large language model based on the matching degrees between the person information of the multiple groups of target videos and the person information included in the multiple target videos includes: Determine, through the large language model, the matching degrees between the person information of each group of target videos and the person information included in the corresponding target video based on the person information of the multiple groups of target videos and the person information included in the multiple target videos; In the case that there are a preset number of matching degrees greater than the preset matching degree among the determined multiple matching degrees, determine that the video to be recognized is a non-original video.

[0013] Another aspect of the embodiments of the present application provides a video source recognition device, and the device includes: An acquisition module for acquiring a video to be recognized; An extraction module for extracting content text and multiple video frames from the video to be recognized; An analysis module for performing text analysis on the content text through a large language model to obtain source information of the content text, where the source information includes a video name; A recognition module for performing face recognition on the multiple video frames to obtain person information included in the multiple video frames; A determination module for determining the video source of the video to be recognized based on the person information recognized by face recognition and the person information included in a target video corresponding to the video name.

[0014] Another aspect of the embodiments of the present application provides a computer device, including: At least one processor; and A memory communicatively connected to the at least one processor; Wherein: the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method as described above.

[0015] Another aspect of the embodiments of the present application provides a computer-readable storage medium, where computer instructions are stored in the computer-readable storage medium, and when the computer instructions are executed by a processor, the method as described above is implemented.

[0016] Another aspect of the embodiments of the present application provides a computer program product, including a computer program, and when the computer program is executed by a processor, the method as described above is implemented.

[0017] The embodiments of the present application adopting the above technical solutions may include the following advantages: First, by analyzing the content text in the video to be recognized through a large language model, accurate identification of the source of the content text is realized automatically. Second, by performing face recognition on the video frames in the video to be recognized, the person information included in the video to be recognized can be identified, that is, which actors are included in the video to be recognized can be identified. Finally, based on the person information and the person information included in a target video corresponding to the video name, the video source of the video to be recognized is judged, that is, it can be determined whether the video to be recognized is an original video or a non-original video (also called an infringing video). Through the above recognition method, it can be quickly and accurately identified whether the video to be recognized is an original video, and thus the occurrence of copyright infringement problems can be effectively prevented. Description of the Drawings

[0018] The accompanying drawings exemplarily illustrate embodiments and form a part of the specification, and are used together with the written description of the specification to explain the exemplary embodiments of the embodiments. The illustrated embodiments are for illustrative purposes only and do not limit the scope of the claims. In all the drawings, the same reference numerals refer to similar but not necessarily identical elements.

[0019] Figure 1 Schematically shows the operating environment diagram of the video source identification method according to Embodiment 1 of the present application; Figure 2 Schematically shows the flowchart of the video source identification method according to Embodiment 1 of the present application; Figure 3 Schematically shows the new flowchart of the video source identification method according to Embodiment 1 of the present application; Figure 4 Schematically shows Figure 2 The sub-step flowchart of step S206 in Figure 5 Schematically shows the refined flowchart of the steps for determining the video source of the to-be-identified video based on the matching degree between the person information of the multi-group of target videos and the person information included in the multiple target videos by the large language model; Figure 6 Schematically shows the block diagram of the video source identification device according to Embodiment 2 of the present application; and Figure 7 Schematically shows the hardware architecture schematic diagram of the computer device according to Embodiment 3 of the present application. Detailed implementation manners

[0020] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0021] It should be noted that the descriptions involving "first", "second", etc. in the embodiments of the present application are only for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of such features. In addition, the technical solutions between various embodiments may be combined with each other, but it must be based on the fact that those of ordinary skill in the art can implement them. When the combination of technical solutions appears to be contradictory or unable to be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the protection scope required by the present application.

[0022] In the description of this application, it should be understood that the numerical labels before the steps do not indicate the order of execution of the steps. They are only used to conveniently describe this application and distinguish each step, and thus should not be construed as a limitation to this application.

[0023] First, provide the term explanations involved in this application: ASR (Automatic Speech Recognition): It is a technology that uses computer technology to automatically analyze and recognize spoken speech and convert it into corresponding text.

[0024] UGC (User Generated Content): It refers to content such as text, pictures, audio, and video created by users published on the Internet in any form, which is a new mode of network information resource creation and organization in the Web2.0 environment.

[0025] LLM (Large Language Model): It refers to a deep learning model trained using a large amount of text data, enabling the model to generate natural language text or understand the meaning of language text. These models can provide in-depth knowledge and language production on various topics by training on a huge dataset. The core idea is to learn the patterns and structures of natural language through large-scale unsupervised training, and to simulate the human language cognition and generation process to a certain extent.

[0026] Face recognition technology: Based on human facial features, for the input face image or video stream, first determine whether there is a face. If there is a face, further give the position, size of each face, and the position information of each main facial organ, and based on this information, further extract the identity features contained in each face and compare them with known faces to identify the identity of each face.

[0027] To facilitate those skilled in the art to understand the technical solutions provided by the embodiments of this application, the related technologies are described below: Traditional UGC platforms generally hire a large number of reviewers to review video content, including the review of copyright issues. However, with the rise of social media and short video platforms, the number of user-generated content (UGC) has increased sharply. Among them, a large amount of UGC contains lines, clips, or character images in the video name, which has led to frequent occurrences of copyright infringement.

[0028] Currently, the recognition of film and television elements in UGC mainly relies on manual review, and its review accuracy and real-time performance are poor.

[0029] To this end, the embodiments of the present application provide a technical solution for video source identification. In this technical solution, first, the content text in the video to be identified is analyzed by a large language model to achieve accurate identification of the source of the content text automatically. Secondly, face recognition is performed on the video frames in the video to be identified, so that the character information included in the video to be identified can be recognized, that is, which actors are included in the video to be identified can be recognized. Finally, based on the character information in the target video corresponding to the video name and the character information included in the video to be identified, the video source of the video to be identified is determined, that is, it can be determined whether the video to be identified is an original video or a non-original video (also called an infringing video). Through the above identification method, it can be quickly and accurately identified whether the video to be identified is an original video, and thus the occurrence of copyright infringement problems can be effectively prevented. See the following for details.

[0030] Finally, for ease of understanding, an exemplary operating environment is provided below.

[0031] As Figure 1 shown, the environmental schematic diagram includes a service platform 2, a network 4, and a client 6, where: The service platform 2 can be composed of a single or multiple computing devices. The multiple computing devices can include virtualized computing instances. The virtualized computing instances can include virtual machines, such as emulations of computer systems, operating systems, servers, etc. The computing devices can load virtual machines based on virtual images and / or other data that define specific software (e.g., operating systems, dedicated applications, servers) for emulation. As the demand for different types of processing services changes, different virtual machines can be loaded and / or terminated on one or more computing devices. A hypervisor can be implemented to manage the use of different virtual machines on the same computing device.

[0032] The service platform 2 can be configured to communicate with the client 6, etc. through the network 4. The network 4 includes various network devices, such as routers, switches, multiplexers, hubs, modems, bridges, repeaters, firewalls, proxy devices, and / or the like. The network 4 can include physical links, such as coaxial cable links, twisted pair cable links, fiber optic links, and combinations thereof, or wireless links, such as cellular links, satellite links, Wi-Fi links, etc.

[0033] The service platform 2 can provide services such as storage, reading, writing, querying, deleting, etc., such as providing video source identification services for the client.

[0034] The client 6 can be an electronic device running an operating system such as Windows, Android™, or iOS, such as a smartphone, tablet device, laptop computer, virtual reality device, gaming device, set-top box, in-vehicle terminal, or smart TV. Based on the above operating systems, various application programs can be run, such as a video source identification program.

[0035] The client 6 can provide / configure a user access page for manipulating the service platform 2 or uploading an object, etc.

[0036] It should be noted that the above devices are exemplary, and in different scenarios or according to different requirements, the number and types of devices can be adjusted.

[0037] The technical solutions of the present application will be introduced below through multiple embodiments. It should be noted that these embodiments can be implemented in many different forms and should not be construed as being limited only to the embodiments described herein.

[0038] Embodiment 1 Figure 2 A flowchart of a video source identification method according to Embodiment 1 of the present application is schematically shown.

[0039] As Figure 2 shown, the video source identification method may include steps S200 to S208, where: Step S200, obtain the video to be identified.

[0040] Step S202, extract content text and multiple video frames from the video to be identified.

[0041] Step S204, perform text analysis on the content text through a large language model to obtain source information of the content text, where the source information includes the video name.

[0042] Step S206, perform face recognition on the multiple video frames to obtain the person information included in the multiple video frames.

[0043] Step S208, determine the video source of the video to be identified based on the person information recognized by face recognition and the person information included in the target video corresponding to the video name.

[0044] The video source identification method provided in this embodiment first analyzes the content text in the video to be identified through a large language model to accurately identify the source of the content text automatically. Secondly, face recognition is performed on the video frames in the video to be identified, so that the character information included in the video to be identified can be recognized, that is, which actors are included in the video to be identified can be recognized. Finally, based on the character information in the target video corresponding to the character information and the video name, the video source of the video to be identified is judged, that is, it can be determined whether the video to be identified is an original video or a non-original video (also called an infringing video). Through the above identification method, it can be quickly and accurately identified whether the video to be identified is an original video, and thus the occurrence of copyright infringement problems can be effectively prevented.

[0045] The following combines Figure 2 to elaborate in detail on each step in steps S200 to S208 and other optional steps.

[0046] Step S200 , obtain the video to be identified.

[0047] The video to be identified refers to the video whose infringement needs to be identified. For example, various self-media videos, advertising videos, etc. can all be used as the video to be identified. The video to be identified can be a video uploaded in real time by the client, or a video uploaded historically but not yet identified.

[0048] Step S202 , extract the content text and multiple video frames from the video to be identified.

[0049] The content text refers to the subtitle text extracted from the video to be identified.

[0050] In one implementation manner, all the video frames extracted from the video to be identified can be used as the multiple video frames extracted.

[0051] In another implementation manner, the video frames in the video to be identified can also be sampled at intervals of a preset number of frames, and all the sampled video frames are used as the multiple video frames. For example, if the video to be identified consists of 300 video frames, when sampling the video frames, sampling is performed every 4 video frames, then the 1st video frame, the 6th video frame, the 11th video frame,..., the 296th video frame will be used as the multiple video frames extracted.

[0052] In the actual application process, the content text can be extracted in various ways. The following provides an exemplary way.

[0053] In an optional embodiment, the extracting the content text from the video to be identified includes: The audio content in the video to be recognized is converted into content text by using automatic speech recognition technology.

[0054] In this embodiment, since there may or may not be subtitle text in the video to be recognized, but it must contain audio content. Therefore, by directly using automatic speech recognition technology to convert the audio content in the video to be recognized into content text, any type of video to be recognized can convert the audio content to obtain content text.

[0055] Another exemplary way to obtain content text is provided below.

[0056] In an alternative embodiment, refer to Figure 3 , the method further includes: Step S300, detecting whether there is subtitle text in the video to be recognized.

[0057] Step S302, when there is subtitle text in the video to be recognized, extracting the subtitle text from the video to be recognized and using the extracted subtitle text as the content text.

[0058] In one implementation, optical character recognition (OCR) technology can be used to directly detect and recognize video frames. When text is recognized in the video frames, it can be determined that there is subtitle text in the video to be recognized.

[0059] In another implementation, a method based on key frame and text position detection is used to detect whether there is subtitle text in the video to be recognized. The detection principle of this method is: Subtitles in a video usually maintain a certain position and form in consecutive frames. By extracting key frames and detecting the text positions therein, it can be determined whether there are subtitles.

[0060] It should be noted that the above subtitle text detection methods are only exemplary, and other detection methods can also be used. For example, with the help of specialized subtitle extraction software, which usually integrates various detection and recognition technologies and can automatically determine whether there are subtitles in a video and extract them.

[0061] In this embodiment, when it is detected that there is subtitle text in the video to be recognized, optical character recognition technology, specialized subtitle extraction software, etc. can be used to extract the subtitle text contained in the video to be recognized from the video to be recognized.

[0062] It should be noted that when the video to be recognized contains a specialized subtitle file, the subtitle text recorded in the subtitle file can also be directly used as the content text.

[0063] In some embodiments, when it is detected that the video to be recognized does not have subtitle text, automatic speech recognition technology is used to convert the audio content in the video to be recognized into content text.

[0064] In this embodiment, before using automatic speech recognition technology to convert the audio content in the video to be recognized into content text, it is first detected whether the video to be recognized has subtitle text. When the video to be recognized has subtitle text, the subtitle text is extracted from the video to be recognized as the content text, and there is no need to use automatic speech recognition technology to convert the audio content in the video to be recognized into content text, which improves the accuracy of content text extraction.

[0065] Step S204 , the large language model performs text analysis on the content text to obtain the source information of the content text, and the source information includes the video name.

[0066] The large language model is obtained by training and learning an existing open-source large language model with relevant data of a large number of videos (including line data, actor information, video names, etc.). The large language model can accurately identify the source of various content texts and will output the source information of the content text, that is, it will output which video or videos the content text comes from. For example, the content text comes from "Video A". Another example is that the content text comes from "Video B in 1986 version" and "Video C in 1987 version". Among them, Video A, Video B, and Video C are the names of the videos.

[0067] Among them, the videos used for training the large language model are the videos that have been uploaded and authorized within the current video platform.

[0068] It should be noted that when there are multiple videos corresponding to the video name, when representing the source information with the video name, the video name needs to include the version number corresponding to the video. Taking the above video name as Video B as an example. Since Video B includes Video B in 1986 version, Video B in 2010 version, Video B in 2011 version, etc., when the large language model outputs the video name, it will add the version number to the video name to distinguish different versions of the video.

[0069] It can be understood that in order to enable the large language model to output more accurate text analysis results, the relevant data of new video names will be fed to the large language model regularly, so that the large language model can learn new knowledge.

[0070] In this embodiment, through text analysis of the content text by the large language model, the source of the content text can be accurately identified.

[0071] Step S206Perform face recognition on the multiple video frames to obtain the person information contained in the multiple video frames.

[0072] The person information refers to the information used to represent the identity of the face, which can be the name, stage name, etc. of the person corresponding to the face.

[0073] In this embodiment, face recognition technology can be used to perform face recognition processing on the video frames to obtain person information. Among them, the face recognition technology can be a face recognition technology based on the template matching method, or a face recognition technology based on deep learning, etc.

[0074] It should be noted that when there are multiple faces in the video frame, face recognition needs to be performed on all the faces to obtain the person information corresponding to each face.

[0075] In the actual application process, the person information contained in the multiple video frames can be obtained in various ways. The following provides an exemplary way.

[0076] In an alternative embodiment, referring to Figure 4 , step S206 may include: Step S400, perform face recognition on the multiple video frames respectively to obtain the person information contained in each video frame.

[0077] Step S402, perform statistical processing on the person information contained in all the video frames to obtain the quantities of multiple different person information.

[0078] Step S404, delete the person information with a quantity lower than the preset quantity, and use the remaining person information as the person information contained in the multiple video frames.

[0079] In some embodiments, after performing face recognition on each video frame, the obtained person information may be the same or different. After performing face recognition on all the video frames, statistical processing can be performed on all the obtained person information to determine how many different person information there are in total, and the quantity of each person information. For example, there are a total of 5 different person information, namely person information a, with a quantity of 10; person information b, with a quantity of 20; person information c, with a quantity of 50; person information d, with a quantity of 2; person information e, with a quantity of 6.

[0080] After obtaining the quantities of each different person information, the person information with a quantity lower than the preset quantity can be deleted, and the remaining person information can be used as the person information contained in the multiple video frames. Among them, the preset quantity can be set according to the actual situation. For example, the preset quantity is 2.

[0081] It should be noted that the remaining person information refers to the person information left after deleting the person information with a quantity lower than a preset quantity from multiple different person information.

[0082] Since there is a certain probability of misrecognition during the face recognition process. The person information with a quantity lower than the preset quantity is very likely the person information obtained after misrecognizing the face. In this embodiment, by deleting the person information with a quantity lower than the preset quantity and using the remaining person information as the person information included in the multiple video frames, the accuracy of the obtained person information can be improved.

[0083] In other embodiments, after obtaining multiple different person information, these different person information can also be used as the person information included in the multiple video frames.

[0084] In other embodiments, when outputting the person information, the confidence level of the person information can also be output. Subsequently, during statistical processing, the person information with a confidence level lower than a preset threshold can also be deleted, the person information with a confidence level higher than the preset threshold can be retained, and these retained person information can all be used as the person information included in the multiple video frames.

[0085] Step S208 , determine the video source of the video to be recognized based on the person information recognized by face recognition and the person information included in the target video corresponding to the video name.

[0086] The person information included in the target video refers to the information used to represent the identity of the person, which can be the name, stage name, etc. of the person.

[0087] The video source is used to determine whether the video to be recognized is an original video or a non - original video.

[0088] It should be noted that an original video refers to a video that is independently conceived, shot, and produced by the creator and does not copy or duplicate the works of others.

[0089] In this embodiment, the person information can be compared with the person information included in the target video corresponding to the video name, and whether the video to be recognized is an original video can be determined according to the comparison result.

[0090] In the actual application process, there are various ways to determine whether the video to be recognized is an original video. The following provides an exemplary way.

[0091] In an alternative embodiment, step S208 may include: Determine the video source of the video to be recognized through the large - language model based on the person information recognized by face recognition and the person information included in the target video corresponding to the video name.

[0092] The matching degree is used to characterize the matching degree between the person information obtained by face recognition and the person information included in the target video corresponding to the video name. The greater the matching degree, the greater the corresponding matching degree value; the lower the matching degree, the smaller the corresponding matching degree value.

[0093] In this embodiment, by leveraging the powerful semantic analysis and semantic understanding capabilities of the large language model, a more accurate infringement determination result can be output, that is, it can be more accurately determined whether the video to be recognized is an original video.

[0094] In other embodiments, it is also possible to directly determine whether the video to be recognized is an original video by comparing the similarity between the person information obtained by face recognition and the person information included in the target video corresponding to the video name.

[0095] In an alternative embodiment, since the content text consists of multiple lines of dialogue. Therefore, the source information of the content text can include multiple video names, that is, the content text is from multiple videos. For example, the content text is from Video A and Video B.

[0096] The method further includes: Combining the person information with the target videos corresponding to multiple video names respectively to obtain multiple sets of person information of the target videos.

[0097] As an example, the person information includes Person 1, Person 2, Person 3, and Person 4. The multiple video names are Video A and Video B respectively. When combining, Person 1, Person 2, Person 3, and Person 4 will be combined with Video A to obtain the person information of the first group of videos. In addition, Person 1, Person 2, Person 3, and Person 4 will also be combined with Video B to obtain the person information of the second group of videos.

[0098] Correspondingly, after obtaining multiple sets of person information of the videos, determining the video source of the video to be recognized by the large language model based on the person information of the face recognition and the person information included in the target video corresponding to the video name includes: Determining the video source of the video to be recognized by the large language model based on the matching degree between the person information of the multiple sets of target videos and the person information included in the multiple target videos.

[0099] In some embodiments, when the large language model determines whether the video to be recognized is an original video, it will first compare and analyze the person information of each set of target videos with the person information included in the corresponding target video respectively to obtain the corresponding matching degree. After obtaining multiple matching degrees, it will comprehensively consider these multiple matching degrees to determine whether the video to be recognized is an original video.

[0100] In this embodiment, when the source information includes multiple video names, the originality of the video to be recognized is determined based on the matching degrees between multiple groups of character information and the character information included in the corresponding target videos, which improves the accuracy of the originality determination of the video.

[0101] In an alternative embodiment, refer to Figure 5 , the video source of the video to be recognized is determined by the large language model based on the matching degrees between the character information of multiple groups of target videos and the character information included in multiple target videos, including: Step S500: The large language model respectively determines the matching degrees between the character information of each group of target videos and the character information included in the corresponding target videos based on the character information of multiple groups of target videos and the character information included in multiple target videos.

[0102] Step S502: When there are a preset number of matching degrees greater than the preset matching degree among the determined multiple matching degrees, it is determined that the video to be recognized is a non-original video.

[0103] In some embodiments, the large language model respectively compares and analyzes the character information of each group of target videos with the character information included in the corresponding target videos, so as to obtain the corresponding matching degrees, that is, multiple matching degrees. After obtaining multiple matching degrees, it is determined whether there are a preset number of matching degrees greater than the preset matching degree among the multiple matching degrees. If there are a preset number of matching degrees greater than the preset matching, it is determined that the video to be recognized is a non-original video. If there are no a preset number of matching degrees greater than the preset matching, it can be determined that the video to be recognized is an original video.

[0104] Among them, the preset number can be set according to the actual situation. For example, the preset number is 2. The preset matching degree is a matching degree threshold set according to the actual situation. For example, the preset matching degree is 0.8.

[0105] In this embodiment, by determining that the video to be recognized is a non-original video only when there are a preset number of matching degrees greater than the preset matching degree among the determined multiple matching degrees, the accuracy of video recognition is improved.

[0106] Embodiment 2 Figure 6 The block diagram of the video source recognition device 600 according to Embodiment 2 of the present application is schematically shown. The device can be divided into one or more program modules. One or more program modules are stored in a storage medium and executed by one or more processors to complete the embodiments of the present application. The program modules referred to in the embodiments of the present application refer to a series of computer program instruction segments that can complete specific functions. The following description will specifically introduce the functions of each program module in this embodiment. AsFigure 6 As shown in Figure 6 , the device 600 may include: an acquisition module 610, an extraction module 620, an analysis module 630, an identification module 640, and a determination module 650, where: The acquisition module 610 is configured to acquire the video to be identified; The extraction module 620 is configured to extract the content text and multiple video frames from the video to be identified; The analysis module 630 is configured to perform text analysis on the content text through a large language model to obtain the source information of the content text, where the source information includes the video name; The identification module 640 is configured to perform face recognition on the multiple video frames to obtain the person information included in the multiple video frames; The determination module 650 is configured to determine the video source of the video to be identified based on the person information recognized by face recognition and the person information included in the target video corresponding to the video name.

[0107] As an optional embodiment, extracting the content text from the video to be identified includes: Converting the audio content in the video to be identified into content text by using automatic speech recognition technology.

[0108] As an optional embodiment, the device 600 is further configured to: Detect whether there is subtitle text in the video to be identified; In the case where the video to be identified has subtitle text, extract the subtitle text from the video to be identified, and use the extracted subtitle text as the content text.

[0109] As an optional embodiment, performing face recognition on the multiple video frames to obtain the person information included in the multiple video frames includes: Performing face recognition on the multiple video frames respectively to obtain the person information included in each video frame; Performing statistical processing on the person information included in all video frames to obtain the quantity of multiple different person information; Deleting the person information with a quantity lower than the preset quantity, and using the remaining person information as the person information included in the multiple video frames.

[0110] As an optional embodiment, determining the video source of the video to be identified based on the person information recognized by face recognition and the person information included in the target video corresponding to the video name includes: Determining the video source of the video to be identified through the large language model based on the person information recognized by face recognition and the person information included in the target video corresponding to the video name.

[0111] As an alternative embodiment, the source information includes multiple video names, and the device 600 is further configured to: Combine the person information with the target videos corresponding to the multiple video names respectively to obtain the person information of multiple groups of target videos; The large language model determines that the video source of the video to be recognized includes: based on the person information recognized by face recognition and the person information included in the target video corresponding to the video name The large language model determines the video source of the video to be recognized based on the matching degree between the person information of the multiple groups of target videos and the person information included in the multiple target videos.

[0112] As an alternative embodiment, the large language model determines the video source of the video to be recognized based on the matching degree between the person information of the multiple groups of target videos and the person information included in the multiple target videos, including: The large language model respectively determines the matching degree between the person information of each group of target videos and the person information included in the corresponding target video based on the person information of the multiple groups of target videos and the person information included in the multiple target videos; In the case that there are a preset number of matching degrees greater than the preset matching degree among the determined multiple matching degrees, it is determined that the video to be recognized is a non-original video.

[0113] Embodiment III Figure 7 FIG. schematically shows a hardware architecture diagram of a computer device 10000 suitable for implementing the video source recognition method according to Embodiment III of the present application. In some embodiments, the computer device 10000 may be a terminal device such as a smart phone, a wearable device, a tablet computer, a personal computer, a vehicle-mounted terminal, a game console, a virtual device, a workbench, a digital assistant, a set-top box, a robot, etc. In other embodiments, the computer device 10000 may be a rack server, a blade server, a tower server, or a cabinet server (including an independent server or a server cluster composed of multiple servers), etc. As Figure 7 shown, the computer device 10000 includes, but is not limited to: a memory 10010, a processor 10020, and a network interface 10030 that can communicate with each other through a system bus. Among them: The memory 10010 includes at least one type of computer-readable storage medium. The readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (such as SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disc, etc. In some embodiments, the memory 10010 may be an internal storage module of the computer device 10000, such as the hard disk or memory of the computer device 10000. In other embodiments, the memory 10010 may also be an external storage device of the computer device 10000, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, Flash Card, etc. equipped on the computer device 10000. Of course, the memory 10010 may also include both the internal storage module and the external storage device of the computer device 10000. In this embodiment, the memory 10010 is generally used to store the operating system and various application software installed on the computer device 10000, such as the program code of the video source recognition method. In addition, the memory 10010 may also be used to temporarily store various data that have been output or will be output.

[0114] In some embodiments, the processor 10020 may be a central processing unit (CPU), controller, microcontroller, microprocessor, or other chip. The processor 10020 is generally used to control the overall operation of the computer device 10000, such as performing control and processing related to data interaction or communication with the computer device 10000. In this embodiment, the processor 10020 is used to run the program code stored in the memory 10010 or process data.

[0115] The network interface 10030 may include a wireless network interface or a wired network interface, which is generally used to establish a communication link between the computer device 10000 and other computer devices. For example, the network interface 10030 is used to connect the computer device 10000 to an external terminal through a network, and establish a data transmission channel and a communication link between the computer device 10000 and the external terminal. The network may be a wireless or wired network such as an enterprise intranet (Intranet), the Internet, the Global System of Mobile communication (GSM for short), Wideband Code Division Multiple Access (WCDMA for short), 4G network, 5G network, Bluetooth, Wi-Fi, etc.

[0116] It should be noted that Figure 7 only the computer device with components 10010 - 10030 is shown, but it should be understood that it is not required to implement all the shown components, and more or fewer components may be implemented alternatively.

[0117] In this embodiment, the video source identification method stored in the memory 10010 may also be divided into one or more program modules and executed by one or more processors (such as the processor 10020) to complete the embodiments of the present application.

[0118] Embodiment 4 The embodiments of the present application further provide a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the video source identification method in the embodiments are implemented.

[0119] In this embodiment, the computer-readable storage medium includes flash memory, hard disks, multimedia cards, card-type memories (e.g., SD or DX memories, etc.), random access memories (RAM), static random access memories (SRAM), read-only memories (ROM), electrically erasable programmable read-only memories (EEPROM), programmable read-only memories (PROM), magnetic memories, magnetic disks, optical discs, etc. In some embodiments, the computer-readable storage medium may be an internal storage unit of a computer device, such as the hard disk or memory of the computer device. In other embodiments, the computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc., equipped on the computer device. Of course, the computer-readable storage medium may also include both the internal storage unit and the external storage device of the computer device. In this embodiment, the computer-readable storage medium is generally used to store the operating system installed on the computer device and various application software, such as the program code of the video source identification method in the embodiment. In addition, the computer-readable storage medium can also be used to temporarily store various data that have been output or will be output.

[0120] Embodiment 5 The embodiment of the present application also provides a computer program product, including a computer program, which when executed by a processor implements the method in the above embodiment.

[0121] Obviously, those skilled in the art should understand that the above-mentioned modules or steps of the embodiments of the present application can be implemented by a general-purpose computer device. They can be concentrated on a single computer device or distributed on a network composed of multiple computer devices. Optionally, they can be implemented by program codes executable by the computer device. Thus, they can be stored in a storage device and executed by the computer device. And in some cases, the steps shown or described can be executed in a different order from here, or they can be separately fabricated into individual integrated circuit modules, or multiple modules or steps among them can be fabricated into a single integrated circuit module to be implemented. In this way, the embodiments of the present application are not limited to any specific combination of hardware and software.

[0122] It should be noted that the above are only the preferred embodiments of the present application, and do not limit the patent protection scope of the present application. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings of the present application, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present application.

Claims

1. A method for identifying a video source, characterized in that, The method includes: Obtaining the video to be recognized; Extracting the content text and multiple video frames from the video to be recognized; Performing text analysis on the content text through a large language model to obtain the source information of the content text, where the source information includes the video name; Performing face recognition on the multiple video frames to obtain the person information included in the multiple video frames; Determining the video source of the video to be recognized based on the person information recognized through face recognition and the person information included in the target video corresponding to the video name.

2. The method according to claim 1, wherein The extracting the content text from the video to be recognized includes: Using automatic speech recognition technology to convert the audio content in the video to be recognized into content text.

3. The method according to claim 2, characterized in that, The method further includes: Detecting whether there is subtitle text in the video to be recognized; When there is subtitle text in the video to be recognized, extracting the subtitle text from the video to be recognized and using the extracted subtitle text as the content text.

4. The method according to claim 1, wherein The performing face recognition on the multiple video frames to obtain the person information included in the multiple video frames includes: Performing face recognition on the multiple video frames respectively to obtain the person information included in each video frame; Performing statistical processing on the person information included in all video frames to obtain the quantities of multiple different person information; Deleting the person information with a quantity lower than the preset quantity and using the remaining person information as the person information included in the multiple video frames.

5. The method according to any one of claims 1 to 4, characterized in that, The determining the video source of the video to be recognized based on the person information recognized through face recognition and the person information included in the target video corresponding to the video name includes: Determining the video source of the video to be recognized through the large language model based on the person information recognized through face recognition and the person information included in the target video corresponding to the video name.

6. The method according to claim 5, characterized in that, The source information includes multiple video names, and the method further includes: Combining the person information with the target videos corresponding to the multiple video names respectively to obtain the person information of multiple groups of target videos; The determining the video source of the video to be recognized through the large language model based on the person information recognized through face recognition and the person information included in the target video corresponding to the video name includes: Determining the video source of the video to be recognized through the large language model based on the matching degrees between the person information of the multiple groups of target videos and the person information included in the multiple target videos.

7. The method according to claim 6, characterized in that, The determining the video source of the video to be recognized through the large language model based on the matching degrees between the person information of the multiple groups of target videos and the person information included in the multiple target videos includes: Determining the matching degrees between the person information of each group of target videos and the person information included in the corresponding target video respectively through the large language model based on the person information of the multiple groups of target videos and the person information included in the multiple target videos; When there are a preset number of matching degrees greater than the preset matching degree among the determined multiple matching degrees, determining that the video to be recognized is a non-original video.

8. A video source identification device, characterized in that, The device includes: An obtaining module, configured to obtain the video to be recognized; An extracting module, configured to extract the content text and multiple video frames from the video to be recognized; An analysis module for performing text analysis on the content text through a large language model to obtain source information of the content text, where the source information includes a video name; An identification module for performing face recognition on the multiple video frames to obtain the person information included in the multiple video frames; A determination module for determining the video source of the to-be-identified video based on the person information recognized by face recognition and the person information included in the target video corresponding to the video name; 9. A computer device, characterized in that, including: At least one processor; and A memory communicatively connected to the at least one processor; wherein: The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions, and when the computer instructions are executed by a processor, the method according to any one of claims 1 to 7 is implemented.

11. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, the steps of the method according to claims 1 to 7 are implemented.

Images