Video violation processing method and device

Abstract

The embodiment of the invention provides a video violation processing method. The method comprises the following steps: acquiring an initial video material uploaded by a target object; inputting the initial video material into a pre-trained video recognition model to obtain a recognition result; wherein under the condition that the identification result is illegal, the identification result comprises illegal text description of the initial video material; under the condition that the recognition result is illegal, inputting the initial video material and the illegal text description into a pre-trained video processing model to obtain a target video material used for replacing the initial video material; wherein the target video material is used for being returned to the target object and / or being transferred to a manual auditing node for auditing. According to the technical scheme provided by the embodiment of the invention, the user side can intelligently modify the video after rejecting the video without perception, so that the video processing efficiency is improved, and the time consumed by manual modification is saved.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of computer technology, and in particular, to a method, apparatus, computer device, computer-readable storage medium, and computer program product for processing video violations. Background Art

[0002] With the rise of social media and online platforms, video promotion has become an important way for brands to communicate with consumers. During the process of video promotion, in order to ensure the safety of the promotional video and the promotion effect after release, the promotion platform will review the promotional video.

[0003] After a promotional video is rejected by the reviewer, the promoter needs to modify the video based on the violation description text returned during the review process, which affects the efficiency of the promotion.

[0004] It should be noted that the above content is not necessarily prior art, nor is it intended to limit the scope of patent protection of this application. Summary of the invention

[0005] The embodiments of the present application provide a video violation processing method, apparatus, computer equipment, computer-readable storage medium, and computer program product to solve or alleviate one or more of the technical problems raised above.

[0006] One aspect of an embodiment of the present application provides a method for processing video violations, the method comprising: Get the initial video material uploaded by the target object; Inputting the initial video material into a pre-trained video recognition model to obtain a recognition result; wherein, if the recognition result is a violation, the recognition result includes a text description of the violation of the initial video material; In the case where the recognition result is a violation, inputting the initial video material and the violation text description into a pre-trained video processing model to obtain a target video material for replacing the initial video material; The target video material is used to be returned to the target object and / or transferred to a manual review node for review.

[0007] Optionally, the method further comprises: When the identification result is no violation, the initial video material is transferred to the manual review node.

[0008] Optionally, the video processing model is used to output a plurality of modified videos according to the violation description text and the initial video material; and obtaining a target video material for replacing the initial video material comprises: Inputting the plurality of modified videos into the video recognition model to screen out a plurality of compliant modified videos from the plurality of modified videos; The target video material is determined according to the plurality of compliant modified videos.

[0009] Optionally, determining the target video material according to the plurality of compliant modified videos includes: Determine the video quality parameters of the compliant modified videos described in each clause; The compliant modified video with the highest video quality parameter among the multiple compliant modified videos is determined as the target video material.

[0010] Optionally, the video quality parameter includes a modification operation similarity; determining the video quality parameter of each of the compliant modified videos includes: Determine the target object to perform common modification operations for manually modifying the video; Determine the similarity of the modification operations of each of the compliant modification videos according to the common modification operations and the actual modification operations; The actual modification operation is the modification operation performed by the video processing model to obtain a corresponding compliant modified video.

[0011] Optionally, the model modification operation includes one or more of the following operations: Editing the content of the initial video material; Editing the initial video material; and The compliant modified video is generated according to the content of the initial video material.

[0012] Optionally, the method further comprises: Transferring the video to be reviewed to a manual review node and receiving a manual review result; wherein the video to be reviewed is the target video material or the initial video material transferred to the manual review node, and when the manual review result is failure to review, the manual review result includes the manual review reason for the video to be reviewed; If the manual review result is that the review is passed, the video to be reviewed is returned to the target object; In the case where the manual review result is a failure to pass the review, the manual review reason of the video to be reviewed and the video to be reviewed are input into the video processing model to obtain a target modified video for replacing the video to be reviewed; Transferring the target modified video stream to the manual review node to receive the manual review result and manual review reason corresponding to the target modified video; In a case where the manual review result corresponding to the target modified video is failure to pass the review, the target modified video and the manual review result corresponding to the target modified video are returned to the target object.

[0013] Another aspect of the embodiments of the present application provides a video violation processing device, the device comprising: an acquisition module, used to acquire an initial video material uploaded by a target object; A first input module is used to input the initial video material into a pre-trained video recognition model to obtain a recognition result; wherein, if the recognition result is a violation, the recognition result includes a description text of the violation of the initial video material; The second input module is used to input the initial video material and the violation description text into a pre-trained video processing model to obtain a target video material for replacing the initial video material when the recognition result is a violation; wherein the target video material is used to return to the target object and / or be transferred to a manual review node for review.

[0014] Another aspect of an embodiment of the present application provides a computer device, including: at least one processor; and a memory communicatively coupled to the at least one processor; Wherein: the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method as described above.

[0015] Another aspect of an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and when the computer instructions are executed by a processor, the method described above is implemented.

[0016] Another aspect of an embodiment of the present application provides a computer program product, including a computer program, which implements the method described above when executed by a processor.

[0017] The above technical solution adopted in the embodiment of the present application may have the following advantages: When the uploaded video material is judged to be illegal, the model is used to process the video material directly. This allows the user to be unaware and intelligently modify the video after it is rejected, thereby improving video processing efficiency and saving time spent on manual modification. BRIEF DESCRIPTION OF THE DRAWINGS The accompanying drawings exemplarily illustrate the embodiments and constitute a part of the specification, and together with the text description of the specification, are used to explain the exemplary implementation of the embodiments. The embodiments shown are for illustrative purposes only and do not limit the scope of the claims. In all drawings, the same reference numerals refer to similar but not necessarily identical elements.

[0018] Figure 1 A diagram schematically shows an operating environment of a video violation processing method according to Embodiment 1 of the present application; Figure 2 A flowchart of a video violation processing method according to Embodiment 1 of the present application is schematically shown; Figure 3 Schematically shows Figure 2 Flow chart of sub-steps of step S204; Figure 4 Schematically shows Figure 3 Flow chart of sub-steps of step S302; Figure 5 Schematically shows Figure 4 Flow chart of sub-steps of step S400; Figure 6 The newly added flow chart of the video violation processing method according to the first embodiment of the present application is schematically shown; Figure 7 The following schematically shows a workflow diagram of a video violation processing method according to Embodiment 1 of the present application; Figure 8 A block diagram schematically shows a video violation processing device according to the second embodiment of the present application; and Fig. 9 The hardware architecture diagram of the computer device according to the third embodiment of the present application is schematically shown. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solutions and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with 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 intended to limit the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present application.

[0020] 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 number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such combination of technical solutions does not exist and is not within the scope of protection required by this application.

[0021] In the description of the present application, it should be understood that the numerical labels before the steps do not indicate the order in which the steps are executed, but are only used to facilitate the description of the present application and to distinguish each step, and therefore should not be understood as a limitation on the present application.

[0022] First, the following terms are explained: CNN: Convolutional Neural Network is a deep learning algorithm used to process image data. It imitates the way biological visual cortex processes images, extracts image features through structures such as convolutional layers, pooling layers, and fully connected layers, and performs classification or recognition tasks.

[0023] RNN: Recurrent Neural Network is an artificial neural network used to process sequence data. It is suitable for processing and predicting time-dependent problems in time series data, such as natural language processing, speech recognition, and time series prediction.

[0024] CTR: Click-Through Rate, an important indicator for measuring the effectiveness of Internet advertising. CTR refers to the ratio of the number of times users click on an ad to the number of times the ad is displayed, expressed as a percentage.

[0025] CVR: Conversion Rate, which refers to the proportion of users who complete specific behaviors (such as purchase, registration, download, etc.) after clicking on an ad.

[0026] eCPM: Effective Cost Per Mille, is an indicator to measure the efficiency of advertising revenue. It indicates the cost paid by advertisers for every 1,000 ad impressions.

[0027] Video script: A text document written for video production that describes the content, structure, and flow of the video.

[0028] Secondly, in order to facilitate those skilled in the art to understand the technical solutions provided in the embodiments of the present application, the relevant technologies are described below: After a video promoting media (such as an advertising creative video) on a promotion platform (such as an advertising delivery platform) is rejected after review, users can only obtain the textual reasons for the rejection provided by the platform. Users need to understand the reasons for the rejection before making targeted modifications to the original advertising creative video. Moreover, compared with modifying other creative content (titles, pictures), modifying a video takes longer and costs users more, which in turn has a negative impact on the advertiser's overall delivery efficiency.

[0029] To this end, the embodiment of the present application provides a technical solution for handling video violations. In this technical solution, AI technology is combined with the advertising creative review and rejection scenario. After the user chooses to enable this function, if the advertising creativity is rejected in the review stage due to non-compliance of the content, the advertising creative video review and rejection AI modification will automatically interpret the review and rejection reasons, and modify the advertising creative video according to the review and rejection reasons and submit it for review again. It can achieve user-unaware modification of advertising creativity after review and rejection, which can greatly improve the overall advertising delivery efficiency of advertisers and save the time spent on manual modification and review of rejected advertising creative videos. See below for details.

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

[0031] like Figure 1 As shown, the operating environment diagram includes: computer equipment 2, network 4 and user terminal 6.

[0032] Computer device 2 can be composed of a single or multiple computing devices, such as 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. The one or more computer devices may include a virtualized computing instance. The computer device can load a virtual machine based on a virtual image and / or other data defining a specific software (e.g., an operating system, a dedicated application, a server) for simulation. As the demand for different types of processing services changes, different virtual machines can be loaded and / or terminated on the one or more computer devices. A variety of models can be run on the computer device 2, such as a video recognition model, a video processing model, and a video quality assessment model, etc.

[0033] The computer device 2 can receive text data (such as initial title text) uploaded by the user terminal 6 through one or more networks 4, and review and modify the text data through various models running on the computer device 2.

[0034] The network 4 may include various network devices, such as routers, switches, multiplexers, hubs, modems, bridges, repeaters, firewalls, proxy devices, and / or the like. The network 4 may include physical links, such as coaxial cable links, twisted pair cable links, fiber optic links, combinations thereof, and / or the like. The network 4 may include wireless links, such as cellular links, satellite links, Wi-Fi links, and / or the like.

[0035] The user terminal 6 can be configured to access the computer device 2. The user terminal 6 may include any type of computer device that can run a game engine, such as: a smart phone, a tablet device, a laptop computer, a smart device (e.g., a smart watch, smart glasses), a virtual reality, a gaming device, a set-top box, a digital streaming device, a car terminal, a smart TV, a TV box, an MP4 (Moving Picture Experts Group Audio Layer IV) player, etc.

[0036] Client 6, can run Windows system, Android (Android TM ) system or iOS system. In addition, the user can also install various applications and program components as needed. Based on the above programs or program components, various functions can be implemented, such as uploading videos, so that the computer device 2 returns a corresponding reply (i.e., the target video material or the illegal text description, etc.).

[0037] The following uses the computer device 2 as the execution subject to introduce the technical solution of the present application through multiple embodiments. It should be noted that these embodiments can be implemented in a variety of different forms and should not be interpreted as being limited to the embodiments described here.

[0038] Embodiment 1 Figure 2 The flowchart of the video violation processing method according to the first embodiment of the present application is schematically shown.

[0039] like Figure 2 As shown, the video violation processing method may include steps S200 to S204, wherein: Step S200: obtaining the initial video material uploaded by the target object.

[0040] Step S202, inputting the initial video material into a pre-trained video recognition model to obtain a recognition result; wherein, when the recognition result is a violation, the recognition result includes a text description of the violation of the initial video material.

[0041] Step S204, when the recognition result is a violation, the initial video material and the violation text description are input into a pre-trained video processing model to obtain a target video material for replacing the initial video material; wherein the target video material is used to return to the target object and / or circulate to a manual review node for review.

[0042] The video violation processing method provided in this embodiment directly processes the video material through the video processing model when evaluating and determining whether the uploaded video material violates the rules. As a result, the user side can be unaware and intelligently modify the video after the video review is rejected, thereby improving the efficiency of video processing and saving the time spent on manual modification.

[0043] The following combination Figure 2 , each step in steps S200~S204 and other optional steps are described in detail.

[0044] Step S200 , get the initial video material uploaded by the target object.

[0045] The initial video material can be a commercial video, promotional video, educational video, entertainment video, etc. When the target object uploads the video material, it can upload it through a web page, upload it through an API interface, synchronize it with a cloud storage service, etc. The target object can upload one or more initial video materials at a time and store the initial video materials in the video library.

[0046] For example, user U (ie, the target object) uploads five initial video materials, namely V1, V2, V3, V4, and V5, to the video library C on the web page.

[0047] Step S202 , input the initial video material into a pre-trained video recognition model to obtain a recognition result; wherein, when the recognition result is a violation, the recognition result includes a text description of the violation of the initial video material.

[0048] The video recognition model can be obtained based on CNN, RNN, etc. The video recognition model can be trained based on the following data: video materials that have passed review (such as manual review or other review mechanisms), script content extracted from reviewed advertising videos, historical records of AI or manual modification of video materials, and historical video materials that have been rejected by review, etc. According to actual needs, the video recognition model can also be obtained by training based on other models or algorithms using other available raw data. In some embodiments, a third-party general or dedicated visual model can also be called through an interface.

[0049] The trained video recognition model is used to review the input video material (such as the initial video material), identify its rejection risk, estimate the probability of review rejection, and locate the category of the rejection reason (i.e., the illegal text description). The rejection reason may involve false propaganda, contain misleading content, or contain uncomfortable content.

[0050] Before the initial video material is input into the video recognition model, the initial video material may also be preprocessed so that the video recognition model can correctly and efficiently understand the video content.

[0051] The video recognition model is used to perform an automated and intelligent preliminary screening of the initial video material. If illegal content is detected, detailed reasons for the violation are provided. This not only significantly improves the review efficiency and ensures the compliance of the video content, but also provides clear guidance for subsequent video content adjustments.

[0052] For example, input V1, V2...V5 into the video recognition model M1, and find that V1 is in violation. The video recognition model gives a text description of the violation: "The video may contain misleading content between 01:05 and 01:08."

[0053] In an optional embodiment, the method further comprises: When the identification result is no violation, the initial video material is transferred to the manual review node.

[0054] The manual review node can analyze and review the creativity, emotion, etc. of the initial video material. When the target object uploads multiple initial video materials to the video library at one time, only some of the initial video materials selected by the target object from the video library can be transferred to the manual review node for review.

[0055] Manual review of initial video material in which the model did not detect any illegal content can further ensure the creativity and emotional expression of the video content, while avoiding potential risks of violations and improving the overall quality of the video material.

[0056] For example, user U selects the initial video material V3 in the video library, and manually reviews the selected V3, and determines that there is incorrect misleading information in V3, and there is a high risk of violation.

[0057] Step S204 , when the recognition result is a violation, the initial video material and the violation text description are input into a pre-trained video processing model to obtain a target video material for replacing the initial video material; wherein the target video material is used to return to the target object and / or flow to a manual review node for review.

[0058] In specific implementation, the video processing model can be obtained based on CNN, RNN, etc. The video processing model can be obtained by training with the following data: various normal video materials, illegal video materials, and video material modification data. According to actual needs, the video processing model can also be obtained by training with other available raw data based on other models or algorithms. In some embodiments, a third-party general or dedicated visual model can also be called through an interface.

[0059] It should be noted that the user can choose whether to use the video violation processing method provided in this embodiment. If the user chooses not to use it, the initial video material uploaded by the user will not be processed by the video violation processing method provided in this embodiment at this stage.

[0060] When the initial video material is identified as violating the regulations, the video processing model can be used to automatically edit the initial video material based on the illegal content. In this way, the user who uploaded the initial video material (i.e. the target object) can be omitted from adjusting the video according to the description of the illegal content, thereby improving the efficiency of the review process and reducing the burden of users having to modify the video multiple times due to not understanding the rules.

[0061] In an optional embodiment, the video processing model is used to output a plurality of modified videos according to the violation description text and the initial video material. Figure 3 As shown, step S204 includes: S300, inputting the plurality of modified videos into the video recognition model to screen out a plurality of compliant modified videos from the plurality of modified videos.

[0062] S302: Determine the target video material according to the plurality of compliant modified videos.

[0063] The modified video obtained by the video processing model will also be reviewed by the video recognition model to ensure that the processed video is fully compliant and improve the security and reliability of the video content.

[0064] For example, the video processing model generates three modified videos V1a, V1b and V1c based on the initial video material V1, and inputs these three modified videos into the video recognition model to screen out two compliant modified videos V1a and V1c without any illegal content.

[0065] In an optional embodiment, if Figure 4 As shown, step S302 may include: S400: Determine the video quality parameters of each of the compliant modified videos.

[0066] S402: Determine the compliant modified video with the highest video quality parameter among the multiple compliant modified videos as the target video material.

[0067] Video quality parameters can be the predicted CTR (click-through rate) or CVR (conversion rate) of the compliant modified video and the eCPM (effective cost per thousand impressions) of different cross-product combinations of the compliant modified video and other related content (such as pictures / videos / landing pages in the advertisement when the video is used as the advertisement title).

[0068] The quality parameters of compliantly modified videos can be evaluated by using a model trained based on CNN, RNN, etc., using a large amount of video materials that have been put into use or simulated use and their corresponding video quality parameters.

[0069] Selecting compliant modified videos with the highest video quality parameters as target video materials and providing high-quality video materials can effectively increase the audience's interest in clicking on the title, thereby increasing the exposure of the content and the promotion effect of the video.

[0070] In an optional embodiment, the video quality parameter includes a modification operation similarity. Figure 5 As shown, step S400 includes: S500: Determine the common modification operation for manually modifying the video performed by the target object.

[0071] S502, determining the modification operation similarity of each of the compliant modified videos according to the common modification operations and the actual modification operations; wherein the actual modification operation is the modification operation performed by the video processing model to obtain the corresponding compliant modified video.

[0072] In the actual evaluation process, the prediction of parameters such as CTR and CVR of compliantly modified videos may be distorted and differ greatly from the user's predicted values ​​of the video delivery effect.

[0073] In view of this, we can also evaluate the similarity between the modification operations performed by the video processing model when modifying the video material and the operations commonly used by users when modifying videos by themselves, and select the compliant modified videos with high similarity as the target video materials. In this way, we can imitate the user's modification habits in the modification method, so that the target video material can better meet the user's expectations and improve the user's satisfaction with the video modification process.

[0074] For example, the compliant modified video V1a is obtained by using mosaics to cover the illegal content, and the compliant modified video V1c is obtained by generating a new video clip based on the script of the initial video material V1 and replacing it. The most commonly used modification method by users is mosaics, and they rarely use clip replacement to modify videos. Therefore, the modification operation similarity of V1a is higher than that of V1c, and V1a is determined to be the target modified video.

[0075] The video processing model can use a variety of operations when modifying video materials. Several exemplary operations are provided below.

[0076] Operation 1: editing the content of the initial video material.

[0077] For example, mosaic or follow-up special effects are used to block part of the content in the video V1, and new audio is generated to replace part of the audio in the initial video material V1, so as to obtain the modified video V1a.

[0078] Operation 2: editing the initial video material.

[0079] For example, the initial video material V1 is edited, segments containing illegal content are deleted, and the order of the video segments is adjusted to obtain a modified video V1b.

[0080] Operation three: generating the compliant modified video according to the content of the initial video material.

[0081] For example, the script of the initial video material V1 is read and understood, a new video segment is generated according to the understood content, and the segment containing the illegal content in the initial video material is replaced with the new video segment to obtain a modified video V1c.

[0082] During the processing process, the video processing model may adopt one or a combination of the above operations, or may adopt other available operations to modify the video material.

[0083] In an optional embodiment, if Figure 6 As shown, the method also includes: S600, the video to be reviewed is transferred to the manual review node, and the manual review result is received; wherein, the video to be reviewed is the target video material or the initial video material transferred to the manual review node, and when the manual review result is failure to review, the manual review result includes the reason for the manual review of the video to be reviewed.

[0084] S602: When the manual review result is passed, the video to be reviewed is returned to the target object.

[0085] S604, when the manual review result is failure to review, the manual review reason of the video to be reviewed and the video to be reviewed are input into the video processing model to obtain a target modified video for replacing the video to be reviewed.

[0086] S606, transferring the target modified video stream to the manual review node to receive the manual review result and manual review reason corresponding to the target modified video.

[0087] S608, when the manual review result corresponding to the target modified video is failure to pass the review, the target modified video and the manual review result corresponding to the target modified video are returned to the target object.

[0088] The video materials are manually reviewed, and the video materials that meet the requirements are returned to the user (i.e. the target object) as the final video. The video materials that fail the first manual review are modified using the video processing model. This can effectively improve the compliance of the video materials and reduce the risk of violations, while ensuring the creativity and attractiveness of the video materials and improving user satisfaction. Videos that still fail to pass manual review after modification will be directly returned, which can simplify the review process and improve the efficiency of the review process.

[0089] For example, user U selects V1a from the video library. After manually reviewing V1a, it is found that there are still illegal contents in it. Therefore, V1a and the corresponding illegal text description are input into the video processing model for modification, and the target modified video V1d is obtained. After manually reviewing V1d, it is found that there are still illegal contents, so V1d and the corresponding illegal text description are returned to user U.

[0090] It should be noted that before entering the manual review, the user may be asked again whether to use the video processing model to modify the initial video material that the video recognition model has determined to have a risk of violation and the user has not allowed to modify it using the video processing model.

[0091] In order to make this application easier to understand, the following Figure 7 An exemplary application is provided. S11, receiving an initial advertisement video (ie, initial video text) input by an advertiser O (ie, a target object).

[0092] S12, determining whether the initial advertising video violates the regulations through a video recognition model.

[0093] If yes, go to step S13, otherwise go to step S21.

[0094] S13, modifying the initial advertisement video through the video processing model to obtain multiple compliant modified videos.

[0095] S14, inputting the plurality of compliant modified videos into a video quality assessment model to obtain video quality parameters and modification operation similarity of each compliant modified video.

[0096] S15, determining the compliant modified video with the highest video quality parameter and modification operation similarity as the target advertising video.

[0097] S16, sending the violation reason of the initial advertisement video and the target advertisement video to the advertiser O for confirmation, and determining whether the confirmation result is acceptance modification.

[0098] If yes, go to step S17, otherwise go to step S22.

[0099] S17, transferring the target advertisement video stream to a manual review node for review, and determining whether the manual review result is approved.

[0100] If no, go to step S18, if yes, go to step S23.

[0101] S18, modifying the target advertisement video through the video processing model to obtain a secondary modified video (i.e., secondary modified text).

[0102] S19, transferring the second modified video stream to the manual review node for review, and determining whether the manual review result is approved.

[0103] If no, proceed to step S20, and if yes, proceed to step S24.

[0104] S20, returning the second modified video and the manual review result corresponding to the second modified video to the advertiser O.

[0105] S21, notifying advertiser O that the initial advertisement video has passed the review.

[0106] S22, transferring the initial advertisement video stream to a manual review node for review, and determining whether the manual review result is approved.

[0107] If no, go to step S18, if yes, go to step S21.

[0108] S23, notifying advertiser O that the target advertising video has passed the review.

[0109] S24, notifying advertiser O that the second modified video has passed the review.

[0110] Embodiment 2 Figure 8The block diagram of the video violation processing device according to the second embodiment 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 are executed by one or more processors to complete the embodiment of the present application. The program module referred to in the embodiment of the present application refers 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. Figure 8 As shown, the device 1000 may include: an acquisition module 1100, a first input module 1200, and a second input module 1300, wherein: The acquisition module 1100 is used to acquire the initial video material uploaded by the target object; The first input module 1200 is used to input the initial video material into a pre-trained video recognition model to obtain a recognition result; wherein, if the recognition result is a violation, the recognition result includes a text description of the violation of the initial video material; The second input module 1300 is used to input the initial video material and the violation text description into a pre-trained video processing model to obtain a target video material for replacing the initial video material when the recognition result is a violation; wherein the target video material is used to return to the target object and / or be transferred to a manual review node for review.

[0111] As an optional embodiment, the apparatus 1000 further includes a flow module, which is used to: When the identification result is no violation, the initial video material is transferred to the manual review node.

[0112] As an optional embodiment, the video processing model is used to output a plurality of modified videos according to the violation description text and the initial video material, and the second input module 1300 is further used to: Inputting the plurality of modified videos into the video recognition model to screen out a plurality of compliant modified videos from the plurality of modified videos; The target video material is determined according to the plurality of compliant modified videos.

[0113] As an optional embodiment, the second input module 1300 is further used for: Determine the video quality parameters of the compliant modified videos described in each clause; The compliant modified video with the highest video quality parameter among the multiple compliant modified videos is determined as the target video material.

[0114] As an optional embodiment, the video quality parameter includes a modification operation similarity, and the second input module 1300 is further used for: Determine the target object to perform common modification operations for manually modifying the video; Determine the similarity of the modification operations of each of the compliant modification videos according to the common modification operations and the actual modification operations; The actual modification operation is the modification operation performed by the video processing model to obtain a corresponding compliant modified video.

[0115] As an optional embodiment, the model modification operation includes one or more of the following operations: Editing the content of the initial video material; Editing the initial video material; and The compliant modified video is generated according to the content of the initial video material.

[0116] As an optional embodiment, the apparatus 1000 further includes a manual review module, which is used to: Transferring the video to be reviewed to a manual review node and receiving a manual review result; wherein the video to be reviewed is the target video material or the initial video material transferred to the manual review node, and when the manual review result is failure to review, the manual review result includes the manual review reason for the video to be reviewed; If the manual review result is that the review is passed, the video to be reviewed is returned to the target object; In the case where the manual review result is a failure to pass the review, the manual review reason of the video to be reviewed and the video to be reviewed are input into the video processing model to obtain a target modified video for replacing the video to be reviewed; Transferring the target modified video stream to the manual review node to receive the manual review result and manual review reason corresponding to the target modified video; In a case where the manual review result corresponding to the target modified video is failure to pass the review, the target modified video and the manual review result corresponding to the target modified video are returned to the target object.

[0117] Embodiment 3 Fig. 9 The schematic diagram of the hardware architecture of a computer device 10000 suitable for implementing the video violation processing method according to the third embodiment of the present application is shown. In some embodiments, the computer device 10000 may be a terminal device such as a smart phone, a tablet computer, a personal computer, a vehicle terminal, a game console, a virtual device, a workbench, 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. Fig. 9As 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, and 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 disk, etc. In some embodiments, the memory 10010 can be an internal storage module of the computer device 10000, such as a hard disk or memory of the computer device 10000. In other embodiments, the memory 10010 can also be an external storage device of the computer device 10000, such as a plug-in hard disk equipped on the computer device 10000, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. Of course, the memory 10010 can also include both the internal storage module of the computer device 10000 and its external storage device. In this embodiment, the memory 10010 is generally used to store the operating system and various application software installed in the computer device 10000, such as the program code of the video violation processing method, etc. In addition, the memory 10010 can also be used to temporarily store various data that have been output or will be output.

[0118] In some embodiments, the processor 10020 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other chips. 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.

[0119] The network interface 10030 may include a wireless network interface or a wired network interface, and the network interface 10030 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 to 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 intranet, the Internet, the Global System of Mobile communication (GSM), Wideband Code Division Multiple Access (WCDMA), 4G network, 5G network, Bluetooth, Wi-Fi, etc.

[0120] It should be pointed out that Fig. 9 Only a computer device having components 10010 - 10030 is shown, but it should be understood that implementation of all of the components shown is not a requirement, and more or fewer components may alternatively be implemented.

[0121] In this embodiment, the video violation processing method stored in the memory 10010 can also be divided into one or more program modules and executed by one or more processors (such as processor 10020) to complete the embodiment of the present application.

[0122] Embodiment 4 An embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored, wherein when the computer program is executed by a processor, the steps of the video violation processing method in the embodiment are implemented.

[0123] In this embodiment, the computer-readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), 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, disk, optical disk, etc. In some embodiments, the computer-readable storage medium can be an internal storage unit of a computer device, such as a hard disk or memory of the computer device. In other embodiments, the computer-readable storage medium can also be an external storage device of a computer device, such as a plug-in hard disk equipped on the computer device, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. Of course, the computer-readable storage medium can also include both the internal storage unit of the computer device and its external storage device. In this embodiment, the computer-readable storage medium is generally used to store an operating system and various application software installed on the computer device, such as the program code of the video violation processing method in the embodiment. In addition, the computer-readable storage medium can also be used to temporarily store various types of data that have been output or are to be output.

[0124] Embodiment 5 An embodiment of the present application also provides a computer program product, including a computer program, which implements the method in the above embodiment when executed by a processor.

[0125] Obviously, those skilled in the art should understand that the modules or steps of the above-mentioned embodiments of the present application can be implemented by general-purpose computer devices, 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 executable program codes of computer devices, so that 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 that herein, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. In this way, the embodiments of the present application are not limited to any specific combination of hardware and software.

[0126] It should be noted that the above are only preferred embodiments of the present application, and the patent protection scope of the present application is not limited thereto. Any equivalent structure or equivalent process transformation made using the contents of the specification and drawings of the present application, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A video violation processing method, characterized in that: The method comprises: Get the initial video material uploaded by the target object; Inputting the initial video material into a pre-trained video recognition model to obtain a recognition result; wherein, if the recognition result is a violation, the recognition result includes a text description of the violation of the initial video material; In the case where the recognition result is a violation, inputting the initial video material and the violation text description into a pre-trained video processing model to obtain a target video material for replacing the initial video material; The target video material is used to be returned to the target object and / or transferred to a manual review node for review.

2. The method according to claim 1, characterized in that The method further comprises: When the identification result is no violation, the initial video material is transferred to the manual review node.

3. The method according to claim 1, characterized in that The video processing model is used to output a plurality of modified videos according to the violation description text and the initial video material; Obtaining a target video material for replacing the initial video material, including: Inputting the plurality of modified videos into the video recognition model to screen out a plurality of compliant modified videos from the plurality of modified videos; The target video material is determined according to the plurality of compliant modified videos.

4. The method according to claim 3, characterized in that Determining the target video material according to the plurality of compliant modified videos includes: Determine the video quality parameters of the compliant modified videos described in each clause; The compliant modified video with the highest video quality parameter among the multiple compliant modified videos is determined as the target video material.

5. The method according to claim 4, characterized in that The video quality parameters include modification operation similarity; Determine the video quality parameters of the compliant modified videos, including: Determine the target object to perform common modification operations for manually modifying the video; Determine the similarity of the modification operations of each of the compliant modification videos according to the common modification operations and the actual modification operations; The actual modification operation is the modification operation performed by the video processing model to obtain a corresponding compliant modified video.

6. The method according to claim 5, characterized in that The model modification operation includes one or more of the following operations: Editing the content of the initial video material; Editing the initial video material; and The compliant modified video is generated according to the content of the initial video material.

7. The method according to claim 1, characterized in that The method further comprises: Transferring the video to be reviewed to a manual review node and receiving a manual review result; wherein the video to be reviewed is the target video material or the initial video material transferred to the manual review node, and when the manual review result is failure to review, the manual review result includes the manual review reason for the video to be reviewed; If the manual review result is that the review is passed, the video to be reviewed is returned to the target object; In the case where the manual review result is a failure to pass the review, the manual review reason of the video to be reviewed and the video to be reviewed are input into the video processing model to obtain a target modified video for replacing the video to be reviewed; Transferring the target modified video stream to the manual review node to receive the manual review result and manual review reason corresponding to the target modified video; In a case where the manual review result corresponding to the target modified video is failure to pass the review, the target modified video and the manual review result corresponding to the target modified video are returned to the target object.

8. A video violation processing device, characterized in that: The device comprises: The acquisition module is used to acquire the initial video material uploaded by the target object; A first input module is used to input the initial video material into a pre-trained video recognition model to obtain a recognition result; wherein, if the recognition result is a violation, the recognition result includes a description text of the violation of the initial video material; The second input module is used to input the initial video material and the violation description text into a pre-trained video processing model to obtain a target video material for replacing the initial video material when the recognition result is a violation; wherein the target video material is used to return to the target object and / or be transferred to a manual review node for review.

9. A computer device, characterized in that: include: at least one processor; and a memory communicatively connected to the at least one processor; wherein: The memory stores instructions that can be executed 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 perform 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 any one of claims 1 to 7 are implemented.

Images