Video push method, device, and storage medium

Abstract

The present disclosure relates to a method, apparatus, and storage medium for pushing videos, and pertains to the field of multimedia technology. The method for pushing videos includes obtaining one or more videos including a video with a photosensitive label and a video without a photosensitive label, where each video has a photosensitive label when determined to be a photosensitive video by a photosensitive video identification model, and pushing, to a browsing user, a video without a photosensitive label among the one or more videos when the browsing user is in a state of filtering photosensitive videos.

Description

Technical Field

[0001] [Cross - reference to Related Applications] This application is based on an application with a Chinese application number of 202210769349.6 and a filing date of July 1, 2022, claims its priority, and all the disclosure content of the Chinese application is incorporated into this application.

[0002] [Technical Field] The present disclosure relates to the field of multimedia technology, and particularly to a method, apparatus, and storage medium for pushing videos.

Background Art

[0003] Photosensitive epilepsy is a disease that causes seizures by visual stimuli such as strong stimuli from light sources and flash stimuli. Currently, some game and animation production companies, when producing scenes with strong visual stimuli, use means such as reducing contrast to minimize the possibility of inducing seizures in photosensitive epilepsy, and usually, a photosensitive epilepsy warning is clearly stated at the beginning of the work.

[0004] With the popularization of the Internet, terminals, and multimedia technology, video creators are no longer limited to professional media companies. Ordinary users can also shoot videos by themselves, post them on the network, and let other users watch them. Some video - based applications predict videos that browsing users may be interested in and push them to the browsing users. Due to the simplicity of video production and posting, many video creators have emerged. Currently, some video creators do not have a deep understanding of photosensitive epilepsy.

Summary of the Invention

[0005] According to a first aspect of some embodiments of the present disclosure, a video pushing method is provided which involves acquiring one or more videos including videos having a photosensitive label and videos not having a photosensitive label, each video having a photosensitive label if determined to be a photosensitive video by a photosensitive video identification model, and pushing to the user one or more videos that do not have a photosensitive label if the user is in a state to filter out photosensitive videos.

[0006] In some embodiments, the push method further includes obtaining feedback from the viewing user on a viewed video, and if the feedback indicates that the viewed video is a photosensitive video, marking the viewed video as a photosensitive video and adding it to a training set for training a photosensitive video identification model.

[0007] In some embodiments, each video has a photosensitive label if a photosensitive effect is added.

[0008] In some embodiments, the push method further includes pushing one or more videos to the viewer if the viewer is not filtering out photosensitive videos.

[0009] In some embodiments, the push method further includes, if the video to be played is a video with a photosensitive indicator, displaying a selection control to determine whether or not to filter photosensitive videos; accepting an action from the user on the selection control; and, if the action indicates filtering photosensitive videos, setting the user to a state where photosensitive videos are filtered.

[0010] In some embodiments, if the video to be played is a video with a photosensitive indicator that the user is viewing for the first time using the same account or on the same device, a control is displayed to select whether or not to filter out the photosensitive video.

[0011] In some embodiments, the selection control is located on a mask layer, and the mask layer is located above the video to be played.

[0012] According to a second aspect of some embodiments of the present invention, a video pushing method is provided, which includes: obtaining a video uploaded by a creator; processing the uploaded video using a photosensitive video identification model to determine whether the uploaded video is a photosensitive video; adding a photosensitive marker to the uploaded video if it is a photosensitive video; and pushing videos without a photosensitive marker to the user if the user is in a state where they filter out photosensitive videos.

[0013] In some embodiments, the push method further includes sending a notification to the creator when it is determined that the uploaded video is a photosensitive video.

[0014] In some embodiments, the push method further includes sending a notification to the creator, receiving feedback submitted by the creator, and, if the feedback indicates that the uploaded video is not a photosensitive video, marking the uploaded video as a non-photosensitive video and adding it to a training set for training a photosensitive video identification model.

[0015] In some embodiments, the photosensitive video discrimination model is a neural network model.

[0016] According to a third aspect of some embodiments of the present disclosure, a video pushing device is provided, which includes a memory and a processor coupled to the memory and configured to perform any one of the aforementioned video pushing methods based on instructions stored in the memory.

[0017] According to a fourth aspect of some embodiments of the present disclosure, a computer-readable storage medium is provided which stores a computer program that, when executed by a processor, implements one of the aforementioned video push methods.

[0018] According to a fifth aspect of some embodiments of the present invention, a computer program product is provided that, when run on a computer, causes the computer to implement one of the aforementioned video push methods.

[0019] According to a sixth aspect of some embodiments of the present invention, a computer program is provided which, when executed by a processor, includes instructions causing the processor to perform one of the aforementioned video push methods.

[0020] Other features and advantages of this disclosure will become apparent from a detailed description of exemplary embodiments of this disclosure, based on the following drawings. [Brief explanation of the drawing]

[0021] To more clearly explain the technical concepts in the embodiments of this disclosure, or in the prior art, the drawings necessary for describing the embodiments or the prior art are briefly described below. Needless to say, the drawings in the following description are only a few embodiments of this disclosure, and those skilled in the art can obtain other drawings based on these without requiring any special ingenuity. [Figure 1] A flowchart of a video push method according to several embodiments of this disclosure is shown. [Figure 2] A flowchart of a video push method according to some other embodiments of the present invention is shown. [Figure 3] A flowchart of a user feedback method according to several embodiments of the present invention is shown. [Figure 4] A flowchart of a creator feedback method according to several embodiments of the present invention is shown. [Figure 5] A flowchart illustrating a user state setting method according to several embodiments of the present invention is shown. [Figure 6] The schematic configuration diagram of a video push device according to some embodiments of the present invention is shown. [Figure 7] The schematic configuration diagram of a video push device according to some other embodiments of the present invention is shown. [Figure 8] The schematic configuration diagram of a video push device according to some other embodiments of the present disclosure is shown. [Figure 9] The schematic configuration diagram of a video push device according to some further other embodiments of the present disclosure is shown.

Modes for Carrying Out the Invention

[0022] Hereinafter, referring to the drawings of the embodiments of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described. However, it is clear that the described embodiments are only a part of the embodiments of the present disclosure, not all of the embodiments. The description of at least one exemplary embodiment below is, in fact, only illustrative and should not be construed as any limitation on the present disclosure and its application or use. All other embodiments obtained by those skilled in the art based on the embodiments of the present disclosure without any special creativity belong to the protection scope of the present disclosure.

[0023] Unless otherwise specified, the relative arrangements, mathematical formulas, and numerical values of the members and steps described in these embodiments do not limit the scope of the present disclosure.

[0024] At the same time, for the convenience of explanation, it should be understood that the dimensions of each part shown in the drawings are not drawn in actual proportional relationships.

[0025] Although the techniques, methods, and devices known to those skilled in the art may not be described in detail, where appropriate, the said techniques, methods, and devices should be regarded as part of the specification.

[0026] In all the examples shown and described herein, any specific values ​​should be interpreted as illustrative only, and not restrictive. Therefore, other examples of the exemplary embodiments may have different values.

[0027] Similar symbols and letters in the following drawings represent similar items. Therefore, once an item is defined in a drawing, it is not necessary to explain it further in subsequent drawings.

[0028] Some video creators have a limited understanding of photosensitive epilepsy and are therefore more likely to post videos that could trigger seizures in individuals with photosensitive epilepsy. Pushing such videos to users sensitive to photosensitive content can easily cause discomfort and may even trigger an attack in those users.

[0029] One of the technical challenges this disclosure aims to address is how to reduce the impact on users sensitive to light-sensitive video and improve security during video viewing in video push notifications.

[0030] Figure 1 shows a flowchart of a video push method according to several embodiments of the present disclosure. These embodiments may be deployed on a server for the video push application. They may also be deployed on other devices as needed, which are not described further in this disclosure.

[0031] As shown in Figure 1, the video push method of this embodiment includes steps S102 to S108.

[0032] Step S102 retrieves the video uploaded by the creator.

[0033] For example, the creator uploads a video stored on their device or on the network via the upload interface provided by the video push application. The video may be, or may not be, created by the creator, or produced by editing video material.

[0034] In step S104, the uploaded video is processed using a photosensitive video identification model to determine whether or not the uploaded video is a photosensitive video.

[0035] In some embodiments, the photosensitive video identification model is a neural network model. For example, an uploaded video may be input to a pre-trained photosensitive video identification model, and a predicted probability calculated by the photosensitive video identification model may be obtained, which represents the probability that the uploaded video is a photosensitive video or not. By comparing the predicted probability with a predetermined threshold, a determination can be made as to whether or not the uploaded video is a photosensitive video.

[0036] In step S106, if the uploaded video is a photosensitive video, a photosensitive marker is added to the uploaded video.

[0037] Photosensitivity indicators may be stored, for example, in the video's metadata or attribute information. Alternatively, the correspondence between video indicators and photosensitivity indicators may be stored using data tables or key-value pairs.

[0038] In some embodiments, a field indicating photosensitivity is added to the information corresponding to the video, and different field values ​​indicate whether the corresponding video has a photosensitive label or not. For example, a new field called photosensitive (photosensitive) is added, and if photosensitive=1, it indicates that the video has a photosensitive label, and if photosensitive=0 or photosensitive=NULL (empty), it indicates that the video does not have a photosensitive label.

[0039] In some embodiments, a video push application provides creators with several preset effects. After uploading a video, creators can add these effects to enhance its visual appeal. Among these effects, the photosensitive effect may be pre-marked. If a photosensitive effect is added to an uploaded video, a photosensitive indicator may be added to it. This ensures accurate marking even if an uploaded video is not inherently photosensitive, but becomes so after a photosensitive effect is added.

[0040] In some embodiments, when an uploaded video is determined to be a photosensitive video, a notification is sent to the creator. This allows the creator to know that the uploaded video is a photosensitive video. The notification may be sent, for example, via station mail of the video push application, or via a route other than the video push application, such as mobile email or email, based on communication information retained by the creator. In addition to the identification result for the video, the notification may further include knowledge about photosensitive videos and photosensitive epilepsy. This allows more creators to take care to avoid relevant visual effects when creating subsequent videos. This improves the overall security of video viewing.

[0041] In step S108, if the user is in a state where they are filtering out light-sensitive videos, videos that do not have a light-sensitive indicator are pushed to the user.

[0042] Users may choose whether or not to filter out photosensitive videos, as needed. If a user has photosensitive videos filtered out, they will not view videos that contain photosensitive indicators.

[0043] It should be noted that the terms "creator" and "viewer" in this disclosure merely refer to their identities in relation to a particular video. In actual application scenarios, a user may be both the creator of a video and a viewer of other videos.

[0044] In some embodiments, the pushed video resides in a video stream, which may be a long-running video stream or a short-running video stream. The user can view different videos pushed to them by switching between them.

[0045] The following describes an example of how to push videos from the perspective of the user viewing the video, with reference to Figure 2.

[0046] Figure 2 shows a flowchart of a video push method according to several other embodiments of the present invention. These embodiments may be deployed on a server for a video push application. They may also be deployed on other devices as needed, which are not described further in this disclosure.

[0047] As shown in Figure 2, the video push method of this embodiment includes steps S202 to S204.

[0048] In step S202, one or more videos are acquired, including videos with a photosensitive label and videos without a photosensitive label, and each video has a photosensitive label if it is determined to be a photosensitive video by the photosensitive video identification model.

[0049] One or more videos retrieved are intended to be pushed to the user. For example, the server calculates videos that the user might be interested in based on a recommendation algorithm and plans to push these videos to the user. Before performing the push operation, the photosensitive video identification model may mark the videos as photosensitive. This helps the server determine whether or not to push all of these videos to the user.

[0050] In step S204, if the user is in a state where they are filtering out light-sensitive videos, one or more videos that do not have a light-sensitive indicator are pushed to the user.

[0051] In some embodiments, if the user is not filtering for photosensitive videos, one or more videos are pushed to the user. That is, the user can view videos with photosensitive labels and videos without photosensitive labels.

[0052] The above embodiment uses a photosensitive video identification model to identify uploaded videos, and if the model determines that a video is photosensitive, it adds a photosensitive marker to the video. This improves the security of video viewing by minimizing adverse effects on users sensitive to photosensitive videos, such as people with photosensitive epilepsy, by pushing videos without the photosensitive marker to the user if the user chooses to filter out photosensitive videos.

[0053] The following example illustrates the training process of a model, using a neural network model as an example for photosensitive video discrimination.

[0054] First, a training set is constructed, which includes marked photosensitive videos and marked non-photosensitive videos, where the marks are, for example, manually screened and marked. Then, the videos in the training set are input into a neural network model to obtain the model's predictions. Finally, the model's loss value is calculated from the discrepancy between the model's predictions and the marks, and the model's parameters are adjusted based on the loss value. This process is repeated multiple times until the convergence condition is met, at which point training is stopped to obtain a photosensitive video recognition model.

[0055] Some embodiments of this disclosure further provide a user feedback mechanism that allows the training set to be expanded by user feedback, thereby further improving the accuracy of the model. Examples of training a model with user feedback are described below with reference to Figures 3 and 4, respectively.

[0056] Figure 3 shows a flowchart of a browsing user feedback method according to several embodiments of the present invention. As shown in Figure 3, the browsing user feedback method of the embodiment includes steps S302 to S306.

[0057] In step S302, videos without photosensitivity indicators are pushed to users who have set their browser to filter out photosensitive videos.

[0058] Step S304 retrieves feedback from users regarding videos they have viewed.

[0059] If a user has set their browser to filter out light-sensitive videos, but still receives light-sensitive videos while browsing, they may submit feedback indicating that the viewed video is a light-sensitive video.

[0060] A video push application may include feedback controls. After a user triggers the feedback controls, they may select a pre-configured feedback type indicating that they have viewed a photosensitive video, or they may manually enter feedback content.

[0061] In step S306, if the feedback indicates that the viewed video is a photosensitive video, the viewed video is marked as a photosensitive video and added to the training set for training the photosensitive video discrimination model.

[0062] In some embodiments, after receiving feedback from viewers, the feedback information may be further reviewed, and if it passes the review, the appropriate video may be added to the training set. This review may be completed manually or automatically according to predetermined review rules or a review algorithm.

[0063] This allows us to mark photosensitive videos that the model failed to identify based on user feedback and add them to the training set. This can help the model further improve its accuracy.

[0064] Figure 4 shows a flowchart of the creator feedback method according to several embodiments of the present invention. As shown in Figure 4, the embodiment Creator The feedback method includes steps S402 to S410.

[0065] In step S402, a photosensitive video identification model is used to process the video uploaded by the creator and determine whether or not the uploaded video is a photosensitive video.

[0066] In step S404, if the uploaded video is a photosensitive video, a photosensitive marker is added to the uploaded video.

[0067] In step S406, if it is determined that the uploaded video is a photosensitive video, a notification is sent to the creator.

[0068] In step S408, after sending a notification to the creator, the feedback submitted by the creator is received. For example, if the creator believes that the uploaded video is not a photosensitive video, they may submit feedback to raise concerns.

[0069] In step S410, if the feedback indicates that the uploaded video is not a photosensitive video, the uploaded video is marked as a non-photosensitive video and added to the training set for training the photosensitive video discrimination model.

[0070] In some embodiments, after receiving feedback from the creator, the feedback information may be further reviewed, and if it passes the review, the appropriate video may be added to the training set. This review may be completed manually or automatically according to predetermined review rules or a review algorithm.

[0071] This allows creators to mark non-photosensitive videos that the model misidentifies based on their feedback and add them to the training set. This can help the model further improve its accuracy.

[0072] The two feedback mechanisms described above utilize user feedback to identify misclassifications in the model, collect more training data, and further optimize the model, thereby improving its predictive accuracy. User security while watching videos is also better ensured.

[0073] An embodiment of the user state setting method of the present invention will be described below with reference to Figure 5.

[0074] Figure 5 shows a flowchart of a user state setting method according to several embodiments of the present invention. As shown in Figure 5, the user state setting method of this embodiment includes steps S502 to S506.

[0075] In step S502, if the video to be played has a photosensitive indicator, a control is displayed to select whether or not to filter out the photosensitive video.

[0076] For example, a video push application retrieves several videos, pushes them to the user sequentially for playback, and checks whether each video to be played has a photosensitivity indicator before playback. If it does not have a photosensitivity indicator, it plays normally. If it does have a photosensitivity indicator, it displays a selection control.

[0077] In some embodiments, to prevent users from frequently making the same settings, the selection control is displayed only when the user is "browsing for the first time". "First time browsing" refers to the first time browsing with the same account or the first time browsing on the same device. That is, if the video to be played is a video with a photosensitive indicator that the browsing user is viewing for the first time with the same account or on the same device, the selection control is displayed to determine whether or not to filter out the photosensitive video.

[0078] In some embodiments, the selection control is located on a mask layer, which is positioned above the video to be played. The transparency of the mask layer is lower than a predetermined value, or it has the effect of blurring the background. This masks the light-sensitive video in the background, allowing the user's attention to be focused on the selection control.

[0079] Step S504 accepts the user's actions on the selection control.

[0080] In step S506, if the operation indicates filtering of photosensitive video, the browsing user is set to a state where photosensitive video is filtered.

[0081] According to the above embodiment, users can set whether or not to filter out photosensitive videos according to their needs, thereby improving viewing security for users sensitive to photosensitive videos, as well as increasing viewing freedom for users who are not sensitive to photosensitive videos. Moreover, since many viewers are also video creators, this setting process also serves to spread knowledge about photosensitive videos.

[0082] An embodiment of the video push device of the present invention will be described below with reference to Figure 6.

[0083] Figure 6 shows a schematic diagram of the configuration of a video push device according to several embodiments of the present invention. As shown in Figure 6, the video push device 60 of the embodiment is an acquisition module configured to acquire one or more videos, including videos with a photosensitive label and videos without a photosensitive label, wherein each video includes an acquisition module 610 having a photosensitive label if it is determined to be a photosensitive video by a photosensitive video identification model, and a push module 620 configured to push to the user one or more videos that do not have a photosensitive label if the user is in a state to filter out photosensitive videos.

[0084] The above embodiment uses a photosensitive video identification model to identify uploaded videos, and if the model determines that a video is photosensitive, it adds a photosensitive marker to the video. This improves the security of video viewing by minimizing the adverse effects of photosensitive videos on users sensitive to them, such as people with photosensitive epilepsy, by pushing videos without the photosensitive marker to the user if the user chooses to filter out photosensitive videos.

[0085] In some embodiments, the push device 60 further includes a feedback module 630 configured to obtain feedback from a browsing user on a viewed video, and if the feedback indicates that the viewed video is a photosensitive video, to mark the viewed video as a photosensitive video and add it to a training set for training a photosensitive video identification model.

[0086] In some embodiments, each video has a photosensitive label if a photosensitive effect is added.

[0087] In some embodiments, the push module 620 is further configured to push one or more videos to the user if the user is not filtering out photosensitive videos.

[0088] In some embodiments, the push device 60 further includes a setting module 640 configured to display a selection control whether or not to filter out photosensitive videos when the video to be played is a video with a photosensitive indicator, accept an operation on the selection control by the user, and if the operation indicates filtering out photosensitive videos, set the user to a state where photosensitive videos are filtered.

[0089] In some embodiments, the configuration module 640 is further configured to display an option control to filter out photosensitive videos if the video to be played is a video with a photosensitive indicator that the viewer is viewing for the first time with the same account or on the same device.

[0090] In some embodiments, the selection control is located on a mask layer, and the mask layer is located above the video to be played.

[0091] An embodiment of the video push device of the present invention will be described below with reference to Figure 7.

[0092] Figure 7 shows a schematic diagram of the configuration of a video push device according to several other embodiments of the present invention. As shown in Figure 7, the video push device 70 of the embodiment includes an upload module 710 configured to retrieve videos uploaded by creators, an identification module 720 configured to process uploaded videos using a photosensitive video identification model to determine whether or not an uploaded video is a photosensitive video, an addition module 730 configured to add a photosensitive marker to an uploaded video if the uploaded video is a photosensitive video, and a push module 740 configured to push videos without a photosensitive marker to a viewer if the viewer is in a state where they are filtering out photosensitive videos.

[0093] The above embodiment identifies uploaded videos using a photosensitive video identification model, and if the model determines that a video is photosensitive, it adds a photosensitive marker to the video. This improves the security of video viewing by minimizing adverse effects on users sensitive to photosensitive videos, such as people with photosensitive epilepsy, by pushing videos without the photosensitive marker to the user if the user chooses to filter out photosensitive videos.

[0094] In some embodiments, the push device 70 further includes a notification module 750 configured to send a notification to the creator when it is determined that the uploaded video is a photosensitive video.

[0095] In some embodiments, the push device 70 further includes a feedback module 760 configured to send a notification to the creator, receive feedback sent by the creator, and if the feedback indicates that the uploaded video is not a photosensitive video, mark the uploaded video as a non-photosensitive video and add it to a training set for training a photosensitive video identification model.

[0096] In some embodiments, the photosensitive video discrimination model is a neural network model.

[0097] If necessary, the push device 60 and the push device 70 may be installed together in the same device, or they may be installed in different devices. This disclosure will not explain this further.

[0098] Figure 8 shows a schematic diagram of the configuration of a video push device according to several other embodiments of the present disclosure. As shown in Figure 8, the video push device 80 of the embodiment includes a memory 810 and a processor 820 coupled to the memory 810 and configured to execute the video push method of any one of the embodiments described above based on instructions stored in the memory 810.

[0099] Here, memory 810 may include, for example, system memory, a fixed non-volatile storage medium, etc. System memory stores, for example, an operating system, applications, a boot loader, and other programs.

[0100] Figure 9 shows a schematic diagram of the configuration of a video push device according to several other embodiments of the present disclosure. As shown in Figure 9, the video push device 90 of the embodiment includes a memory 910 and a processor 920, and may further include an input / output interface 930, a network interface 940, a storage interface 950, and the like. These interfaces 930, 940, 950, and the memory 910 and processor 920 may be connected, for example, via a bus 980. Here, the input / output interface 930 provides a connection interface to input / output devices such as a display, mouse, keyboard, and touch panel. The network interface 940 provides a connection interface to various devices connected to a network. The storage interface 950 provides a connection interface to external storage devices such as SD cards and U disks.

[0101] Embodiments of the present disclosure further provide a computer-readable storage medium characterized in that it stores a computer program that, when executed by a processor, implements one of the aforementioned video push methods.

[0102] As those skilled in the art will see, embodiments of the present disclosure may be provided as methods, systems, or computer program products. Accordingly, the present disclosure may take the form of complete hardware embodiments, complete software embodiments, or embodiments combining software and hardware. The present disclosure may also take the form of a computer program product implemented on one or more computer-usable non-temporary storage media (including, but not limited to, magnetic disk memory, CD-ROM, optical memory, etc.) containing computer-usable program code.

[0103] This disclosure has been described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of this disclosure. It should be understood that each flow and / or block in a flowchart and / or block diagram, and combinations of flows and / or blocks in a flowchart and / or block diagram, may be implemented by computer program instructions. By providing these computer program instructions to the processor of a general-purpose computer, a dedicated computer, an embedded processor, or other programmable data processing device, a machine can be generated such that instructions executed by the processor of the computer or other programmable data processing device generate a device for implementing one or more flows in a flowchart and / or one or more blocks in a block diagram.

[0104] These computer program instructions may be stored in computer-readable memory that can guide a computer or other programmable data processing device to operate in a particular manner. This causes the instructions stored in this computer-readable memory to generate a product that includes an instruction unit that implements a function specified in one or more flows of a flowchart and / or one or more blocks of a block diagram.

[0105] These computer program instructions may be loaded into a computer or other programmable data processing device. This causes the computer or other programmable device to execute a series of operational steps, generating the processing to be performed by the computer, thereby providing the instructions executed by the computer or other programmable device for performing the functions specified in one or more flows of a flowchart and / or one or more blocks of a block diagram.

[0106] The foregoing describes only preferred embodiments of the Disclosure and is not intended to limit the Disclosure. Any amendments, equivalent substitutions, improvements, etc., made within the spirit and principles of the Disclosure should be included within the scope of the claims of the Disclosure.

Claims

1. Acquire one or more videos, including videos with photosensitivity labels and videos without photosensitivity labels, wherein each video has a photosensitivity label if determined to be a photosensitive video by a photosensitive video identification model, the photosensitive video identification model is trained using a training set, and if it is determined that a video uploaded by the creator is a photosensitive video, and a notification is sent to the creator, and feedback sent by the creator is received, in response to the feedback indicating that the uploaded video is not a photosensitive video, the uploaded video is marked as a non-photosensitive video and added to the training set. A video pushing method, which includes pushing to a user one or more videos that do not have a photosensitive indicator when the user is in a state where they are filtering out photosensitive videos.

2. To obtain feedback from the aforementioned viewing user regarding the videos they have viewed, The push method according to claim 1, further comprising marking the viewed video as a photosensitive video and adding it to the training set if the feedback indicates that the viewed video is a photosensitive video.

3. The push method according to claim 1, wherein each video has a photosensitive label if a photosensitive effect is added.

4. The push method according to claim 1, further comprising pushing one or more videos to the user if the user is not filtering out photosensitive videos.

5. If the video to be played contains a photosensitive content indicator, display a control to select whether or not to filter out the photosensitive video. Accepting operations on the selection control by the browsing user, The push method according to claim 1, further comprising setting the browsing user to a state that filters out photosensitive video if the operation indicates filtering out photosensitive video.

6. The push method according to claim 5, wherein if the video to be played is a video with a photosensitive indicator that the viewing user is viewing for the first time with the same account or on the same device, a control is displayed to select whether or not to filter out photosensitive videos.

7. The push method according to claim 5, wherein the selection control is located on a mask layer, and the mask layer is located on top of the video to be played.

8. Retrieving the video uploaded by the creator, The process involves processing uploaded videos using a photosensitive video identification model to determine whether or not the uploaded videos are photosensitive videos, wherein the photosensitive video identification model is trained using a training set. If the uploaded video is a photosensitive video, a photosensitive marker is added to the uploaded video. If a user is in a state where they are filtering out light-sensitive videos, the system will push videos that do not have a light-sensitive indicator to the user. If a video uploaded by a creator is determined to be a photosensitive video, a notification will be sent to the creator. Receiving the feedback sent by the aforementioned creator, A method for pushing a video, comprising: marking the uploaded video as a non-photosensitive video and adding it to the training set in response to the feedback indicating that the uploaded video is not a photosensitive video.

9. After sending a notification to the aforementioned creator, the system receives the feedback sent by the aforementioned creator. The push method according to claim 8, further comprising marking the uploaded video as a non-photosensitive video and adding it to a training set for training the photosensitive video identification model if the feedback indicates that the uploaded video is not a photosensitive video.

10. The push method according to claim 8, wherein the photosensitive video identification model is a neural network model.

11. Memory and A video pushing device comprising a processor coupled to the memory and configured to perform the video pushing method according to any one of claims 1 to 10 based on instructions stored in the memory.

12. A computer-readable storage medium storing a computer program that, when executed by a processor, enables the video pushing method described in any one of claims 1 to 10.

13. A computer program that, when executed by a processor, includes instructions causing the processor to perform the video push method described in any one of claims 1 to 10.

Images