An AI video creation platform collaborative management method based on deep learning

Abstract

The application discloses an AI video creation platform collaborative management method based on deep learning, belongs to the technical field of computer data processing, deep learning and film and television creation, and realizes intelligent configuration and dynamic adjustment of collaborative permissions through a deep learning model of a CNN+LSTM architecture, establishes a cloud collaborative creation canvas to support real-time online operation of multiple persons, realizes one-key import of a selected lens into a shot list and real-time synchronization of a comment modification state, realizes real-time editing of multiple terminals through a shareable link with configurable permissions, combines AI assistance to complete creation progress reminding and modification suggestion optimization, and realizes whole-process tracking of modification traces and cloud AES256 encryption multi-node backup. The application solves the problems of the existing platform, such as asynchronous collaboration, poor cross-terminal compatibility and low management efficiency, the collaborative delay of the multiple-person canvas is less than 0.5 seconds, the shot comment efficiency is improved, and the application is suitable for industrialized collaborative needs of film and television level video creation.

Description

Technical Field

[0001] This invention relates to the fields of computer data processing, deep learning, and film and television creation technology, and in particular to a collaborative management method for an AI video creation platform based on deep learning. Background Technology

[0002] In the industrialized video creation process of the professional film and television industry, the core links of the entire creative system are the main creative team's early planning, creative polishing, division of labor, scheme design, annotation and iterative modification, and standardized multi-person collaborative creation. Cross-platform editing is a key capability to improve creative flexibility and adapt to diverse creative scenarios. However, the collaborative management technology of mainstream video creation platforms still has many shortcomings and deficiencies, making it difficult to adapt to the creative needs of the professional film and television industry. (1) Multi-person collaborative creation is mostly local offline collaboration. Even if online collaboration is supported, it is impossible to achieve real-time synchronous operation of multiple people on the same creation canvas. The operation trajectory and modified content cannot be shared in real time, resulting in low collaboration efficiency. (2) After the script, pictures and video footage are imported into the storyboard list, the annotations and modification opinions are mostly transmitted offline. The annotation function in the platform only supports single user editing. The status before and after modification cannot be displayed in real time. Managers and production staff cannot view the modification progress at the same time, which can easily lead to communication discrepancies. (3) Poor cross-platform editing compatibility. The sharing of created content between devices such as computers, tablets, and mobile phones mostly relies on file transfer. The content shared through links is mostly in read-only mode, which cannot achieve cross-platform real-time editing. Furthermore, there are problems such as delays and data loss in multi-platform data synchronization. (4) Collaborative management lacks AI assistance, making it impossible to dynamically adjust permissions and provide intelligent reminders for modification opinions based on the user's creative role and work progress. At the same time, modification traces cannot be traced throughout the entire process, making it difficult to troubleshoot problems later.

[0003] Furthermore, while some existing video creation platforms offer basic collaboration and annotation functions, they lack customized development for the professional needs of film and television production, such as storyboard design and shot editing. Importing selected shots into the storyboard list is cumbersome, and annotation formatting is inconsistent, failing to meet the professional collaboration requirements of film and television-level video creation. In summary, existing technologies are ill-suited to the efficient, professional, and flexible collaboration needs of film and television-level video creation. There is an urgent need for a method that enables real-time multi-user canvas collaboration, synchronized storyboard annotation status, seamless cross-platform editing, and intelligent collaborative management through deep learning AI. Summary of the Invention

[0004] The purpose of this invention is to solve the problems existing in the collaborative management technology of existing video creation platforms, such as asynchronous collaboration among multiple users, low efficiency in modifying scripts, images and video storyboard annotations, poor cross-platform editing compatibility, lack of AI assistance, and inability to trace modification traces. This invention provides a deep learning-based AI video creation platform collaborative management method to improve the collaborative management efficiency and professionalism of video creation platforms and adapt to the industrial collaborative needs of film and television-level video creation.

[0005] To achieve the above objectives, this invention provides a collaborative management method for an AI video creation platform based on deep learning, comprising the following steps: Step S1: Based on artificial intelligence deep learning, analyze the user's created content through AI models and assign refined initial collaborative permissions to different users; Step S2: Extract features and perform structured classification on the uploaded scripts, images, and video footage. Supports one-click import of selected shots into a custom storyboard list, realizing structured storage and visual display of shot footage. Step S3: Establish a cloud-based collaborative creation canvas. Authorized users can simultaneously perform creative operations on the same cloud-based collaborative creation canvas. Through a real-time data incremental update algorithm, the operation trajectory and creative modifications of each user are synchronized to the canvas interface of all online users in real time. Step S4: Allows users to annotate and edit any shot in the storyboard list, records the state of the storyboard before and after modification in real time, and synchronizes the annotations and modification status to the storyboard list interface of all authorized users in real time. Step S5: Generate a unique sharing link for collaborative content and storyboard list, supporting permission configuration for the link. Users can use the link to achieve cross-device real-time editing across multiple terminal devices, and the content of cross-device operations is synchronized with the cloud in real time. Step S6: Monitor the collaborative creation progress in real time through a deep learning AI model, provide intelligent reminders for incomplete modification tasks and comments to be confirmed, and intelligently optimize the modification opinions of storyboard annotations based on industry standards for film and television creation. Step S7: Structure and store all users' operation trajectories, storyboard annotations, and modification records, generate a unique identifier for each modification, support multi-dimensional query of modification traces, and realize full-process traceability and restoration of modified content; Step S8: Encrypt all collaborative creation data in the cloud and perform multi-node off-site backups.

[0006] Preferably, in step S1, the deep learning AI model is a user behavior recognition model, which adopts a CNN+LSTM architecture and has been pre-trained using a large amount of user behavior data for video creation; the analyzed user historical creation behavior includes editing frequency, annotation type and management operation; the assigned initial collaborative permissions include editing rights, annotation rights, viewing rights and management rights.

[0007] Preferably, in step S2, the uploaded scripts, images, and video footage are subjected to CNN-based visual feature extraction, and are structurally categorized according to shot type, shooting angle, and content; the storyboard list supports drag-and-drop sorting of shots, categorization by scene, and keyword filtering.

[0008] Preferably, in step S3, the WebSocket real-time communication protocol is used to realize the real-time data transmission of the cloud collaborative creation canvas; the real-time data incremental update algorithm only transmits the modified partial data instead of the full data; and the multi-person collaborative synchronization delay of the cloud collaborative creation canvas is less than 0.5 seconds.

[0009] Preferably, in step S4, the storyboard annotations are displayed in red next to the storyboard screen, and the completion status is displayed in green simultaneously; all authorized users can view the storyboard modification progress, modification content, and annotation comments in real time.

[0010] Preferably, in step S5, the shared link supports configuring editing or viewing rights, and can be shared across multiple terminal devices such as computers, tablets, and mobile phones; the cross-device editing data synchronization delay is less than 1 second.

[0011] Preferably, in step S6, the intelligent reminder methods include platform messages and SMS reminders; the intelligent optimization of storyboard annotation modification opinions is to transform vague modification opinions into specific modification suggestions that conform to the standards of the film and television creation industry.

[0012] Preferably, in step S7, the unique identifier generated for each modification is a time-user identifier; the stored modification information includes the modification content, modification time, operator, and state before and after modification; and it supports querying modification traces by time, user, and scene number.

[0013] Preferably, in step S8, the collaborative creation data is encrypted and stored in the cloud using the AES256 encryption algorithm; multi-node off-site backups are performed on Alibaba Cloud, Tencent Cloud, and Huawei Cloud, with incremental backups performed every 5 minutes.

[0014] The preferred option is integrated into the Xcine film and television-grade AI creation platform, targeting professional film and television production teams, enabling multi-person collaborative creation, storyboard annotation and modification, and cross-platform editing for directors, cinematographers, storyboard artists, production staff, editors, and production managers; the operation of importing selected shots into the storyboard list takes less than 1 second per shot.

[0015] Therefore, the above-mentioned collaborative management method for an AI video creation platform based on deep learning, as described in this invention, has the following advantages: (1) Based on the deep learning AI model, the intelligent configuration and dynamic adjustment of collaborative permissions are realized, which avoids the collaborative chaos caused by unreasonable permission allocation and improves the professionalism of collaborative management. (2) A cloud collaborative creation canvas was established, and the WebSocket protocol and incremental update algorithm were combined to realize real-time online collaborative creation by multiple people with a synchronization delay of less than 0.5 seconds, which greatly improved the efficiency of multi-person collaboration. (3) Supports one-click import of selected shots into the storyboard list, with an operation time of less than 1 second per shot, and realizes real-time synchronization of storyboard annotation and modification status, eliminating communication deviations and improving the efficiency of storyboard annotation and modification by more than 90% compared with traditional methods. (4) Cross-platform real-time editing is achieved by using configurable permission-based link sharing, with cross-platform synchronization delay of less than 1 second, which solves the problems of poor compatibility and data synchronization delay in traditional cross-platform editing; (5) Combining AI assistance to achieve intelligent reminders of creative progress and intelligent optimization of modification suggestions provides professional references for creators and further improves the efficiency of collaborative creation; (6) Enable full-process traceability and version management of modification traces, which facilitates later problem investigation and adapts to the industrial process requirements of film and television-level video creation; (7) AES256 encryption and multi-node off-site backup are used to ensure the security and integrity of collaborative creation data and avoid creative loss caused by data loss; (8) It is specially developed for film and television-level video creation, which fits the actual process of film and television creation, greatly reduces the professional threshold of film and television creation collaboration, and improves the efficiency of collaborative management by more than 85% compared with traditional platforms.

[0016] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0017] Figure 1 This is an overall flowchart of a collaborative management method for an AI video creation platform based on deep learning, as described in an embodiment of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0019] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0020] Example like Figure 1 As shown, this embodiment proposes a collaborative management method for an AI video creation platform based on deep learning, integrated into the Xcine film and television-grade AI creation platform. It is designed for professional film and television video creation teams, enabling collaborative creation, storyboard annotation and modification, and cross-platform editing among directors, art directors, cinematographers, storyboard artists, production staff, editors, and production managers. The specific implementation steps are as follows: Step 1: Smart Configuration of Collaborative Permissions: Creative teams can log in to the Xcine platform via computer or mobile device, upload creative materials, and set role tags. Roles include director, art director, cinematographer, storyboard artist, producer, editor, and production manager. The platform's AI collaborative management module incorporates a user behavior recognition model based on a CNN+LSTM architecture. This model is pre-trained using massive amounts of film and television creative behavior data and intelligently analyzes user roles and historical operation data (editing frequency, annotation types, project management actions, etc.) to configure tiered initial collaborative permissions. Director: Full authority, including content editing, manuscript annotation, project management, and member permission allocation; Artwork: Right to edit art materials + right to annotate + right to access project information; Photography: Right to upload and edit footage, add annotations, and review project details; Storyboard: The right to edit and annotate the storyboard script; Production rights: final editing rights + annotation rights + project access rights; Editing: Right to edit footage + right to view project materials; Producer: Right to annotate content + right to view the entire project + right to monitor project progress.

[0021] After the storyboard artists complete all the rectification tasks assigned by the director, the AI ​​model automatically adds and confirms the production permissions for the shots, realizing dynamic upgrading and downgrading of permissions.

[0022] This method centers on the standardized production process of the film and television industry. The director oversees the overall creative style, narrative logic, and camera work standards, utilizing full platform access to complete project initiation, creative requirement input, and finalization of the overall creative plan. Production managers, with project progress monitoring and full project viewing rights, manage the entire project timeline, control the creative progress, coordinate inter-positional processes, and record work milestones for each role to avoid gaps in workflow and delays. Art directors, with their exclusive rights to edit and annotate art materials, independently complete the overall art direction for the project, including core art elements such as scene style, color scheme, character design, and prop visual standards. They generate a complete set of art direction materials and upload them to the platform, while also annotating and optimizing the visual content at each stage. Cinematographers, with their rights to upload, edit, and annotate footage, combine the director's creative requirements and art direction standards to develop shooting plans, clarify shot arrangement rules and shooting parameter standards, complete pre-production shot creation guidance, and upload shooting specifications and reference shot materials for the entire team to review and reference. The creative results from each position are synchronized to the platform's material library in real time, providing a standardized basis for subsequent storyboard creation, material production, and final editing.

[0023] Step Two: Import Selected Shots into the Storyboard List with One Click: After the cinematographers completed the on-location shooting according to the pre-established shooting standards, they uploaded all the footage to the Xcine platform in batches, leveraging their editing rights. The platform, equipped with a CNN visual feature extraction algorithm, automatically performs frame-level feature analysis on all footage, intelligently classifying and archiving it according to shot size, shooting angle, composition, and content scene, forming a standardized footage library. Editors, with their shot editing rights and project review rights, browse the categorized footage library and, based on the director's confirmed creative script, cinematography standards, and art direction, select high-quality shots. Using the platform's one-click import function, they add the selected compliant shots in batches to a custom storyboard list. The storyboard list supports visual preview, drag-and-drop shot sorting, scene categorization management, and precise keyword filtering. All editing operations are synchronized to the cloud in real time. The director can review the shot selection results with full permissions, and the producer can simultaneously monitor the footage collection progress to ensure that the footage conforms to the project's creative standards.

[0024] Step 3: Real-time collaborative creation on a multi-canvas: The director creates a dedicated cloud-based collaborative creation canvas through the platform, synchronizing the finalized creative plan, art direction standards, cinematography specifications, and a list of selected shots to the canvas. Corresponding permissions are granted as needed, allowing storyboard artists, art directors, cinematographers, editors, and producers to access and collaborate online. The platform uses the WebSocket real-time communication protocol and incremental data update algorithm, synchronizing only partial changes made by users, abandoning the traditional full data transmission mode, and significantly reducing transmission bandwidth and latency. Storyboard artists, relying on their editing and annotation rights, strictly adhere to the director's narrative requirements, art direction visual standards, and cinematography specifications, completing the drawing, layout, and optimization of the entire storyboard in the collaborative canvas. During the creation process, art directors can annotate and correct the visual style and color scheme of the storyboard shots, while cinematographers can optimize and annotate the shot arrangement and shooting logic of the storyboard. All modification trajectories and annotations are incrementally synchronized in real time, with a synchronization delay of ≤0.3 seconds across the canvas interfaces of each user, enabling multi-person parallel real-time collaborative creation and completely solving the problems of communication delays and modification conflicts inherent in traditional offline storyboard creation.

[0025] After the storyboard artist completes the entire storyboard, it is submitted to the director for review and finalization. The director, with full authority, finalizes the storyboard and locks in the compliant storyboard solution. Production staff, relying on their editing, annotation, and project review rights, use the director's finalized storyboard as the sole basis, combining pre-production art direction materials and cinematography reference shots. Leveraging the platform's AI capabilities, they generate corresponding scene images, composite video clips, and optimize the material for each scene. During production, the producer monitors the material production progress in real-time and updates project timelines. The director can review the production results and provide optimization suggestions at any time. Art direction and cinematography staff can review the visual style and shot compatibility of the materials and annotate and modify them. Production staff iterate and optimize the composite material based on the compliant annotations from each role, ensuring that all generated content fully conforms to the overall project creative standards.

[0026] Step 4: Synchronize storyboard annotations and modification status: For storyboards and AI-generated footage within the canvas, each role can initiate compliant annotation and modification operations according to their own permissions: production managers can offer optimization suggestions from a project implementation perspective, art directors and cinematographers can correct visual and shot-related issues from a professional perspective, and directors can issue final creative control and modification instructions. The platform's storyboard annotation and modification synchronization module automatically retains the original state of all content before modification, displaying annotation content and modification dimensions with differentiated identifiers; after production staff and storyboard artists complete their corresponding rectification tasks, the platform updates the modification status in real time and synchronizes it to all authorized user interfaces. After all storyboards and footage are reviewed and finalized, editors, relying on their shot editing rights, strictly follow the shot logic and narrative rhythm confirmed by the director to edit, splice, optimize transitions, and adjust the rhythm of live-action footage and AI-generated video clips, completing the initial cut editing work, under the supervision of the director and the control of the production progress throughout the process.

[0027] Step 5: Cross-platform link sharing and real-time editing: To adapt to mobile office and remote collaboration scenarios, the platform supports cross-platform link sharing and real-time editing. Directors, producers, and other staff can generate unique, permission-controlled sharing links through the platform, customize link viewing, annotation, and editing permissions, and achieve content interoperability and real-time operation across computers, mobile phones, and tablets. When on the go, directors can view the entire progress of storyboard creation, material production, and final editing on their mobile devices, add annotations and modifications at any time, and the mobile operation data is synchronized to the cloud canvas and computer in real time. Staff can receive and implement modifications instantly, with cross-platform data synchronization latency of ≤1 second.

[0028] Step Six: AI-Assisted Collaborative Management The AI-assisted collaborative management module monitors the entire creative process, automatically sending intelligent reminders to the relevant personnel and director for overdue storyboarding, material production, and editing tasks. Based on film and television industry standards, the AI ​​model professionally optimizes annotations for each role, outputting standardized modification suggestions to assist each role in efficiently completing creative iterations. Simultaneously, it continuously and dynamically adjusts role permissions based on user task completion and operational behavior data, enabling adaptive permission upgrades and downgrades.

[0029] Step Seven: Trace the entire process of modification traces: The full-track modification module generates a unique time-user identifier for all actions performed by all roles throughout the project lifecycle, including storyboard creation, material generation, annotation modifications, editing adjustments, and permission changes. It then stores structured data such as operator, operation time, modified content, pre- and post-modification status, and permission change records. During project debriefing, issue tracing, or version iteration, directors and managers can precisely retrieve modification traces at any stage by storyboard number, time point, and operator role. It also supports one-click restoration of any historical version, achieving full-process traceability, reviewability, and controllability.

[0030] Step 8: Collaborative Data Cloud Storage and Backup: The Xcine platform uses the AES256 encryption algorithm to encrypt and store all core data, including collaborative canvas data, storyboard data, source files, annotation records, editing projects, and permission logs, in the cloud. It also performs multi-node off-site incremental backups on Alibaba Cloud, Tencent Cloud, and Huawei Cloud, automatically performing an incremental backup every 5 minutes to completely avoid the risk of data loss and leakage, and ensure the security and integrity of data throughout the entire film and television creation process.

[0031] In this embodiment, the synchronization latency of multi-person canvas collaboration in the method of the present invention is less than 0.5 seconds, the synchronization latency of cross-platform editing data is less than 1 second, the operation time for importing selected shots into the storyboard list is less than 1 second per shot, the efficiency of storyboard annotation modification is more than 90% higher than that of traditional offline methods, and the efficiency of collaborative management is more than 85% higher than that of traditional video creation platforms. Through testing on 50 film and television short drama creation teams, the user satisfaction rate of this method reached 99%, effectively solving the core pain points in film and television creation collaboration such as unclear job responsibilities, disconnected processes, delayed collaboration, rigid permissions, and data insecurity.

[0032] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A collaborative management method for an AI video creation platform based on deep learning, characterized in that, Includes the following steps: Step S1: Based on artificial intelligence deep learning, analyze the user's created content through AI models and assign refined initial collaborative permissions to different users; Step S2: Extract features and perform structured classification on the uploaded scripts, images, and video footage. Supports one-click import of selected shots into a custom storyboard list, realizing structured storage and visual display of shot footage. Step S3: Establish a cloud-based collaborative creation canvas. Authorized users can simultaneously perform creative operations on the same cloud-based collaborative creation canvas. Through a real-time data incremental update algorithm, the operation trajectory and creative modifications of each user are synchronized to the canvas interface of all online users in real time. Step S4: Allows users to annotate and edit any shot in the storyboard list, records the state of the storyboard before and after modification in real time, and synchronizes the annotations and modification status to the storyboard list interface of all authorized users in real time. Step S5: Generate a unique sharing link for collaborative content and storyboard list, supporting permission configuration for the link. Users can use the link to achieve cross-device real-time editing across multiple terminal devices, and the content of cross-device operations is synchronized with the cloud in real time. Step S6: Monitor the collaborative creation progress in real time through a deep learning AI model, provide intelligent reminders for incomplete modification tasks and comments to be confirmed, and intelligently optimize the modification opinions of storyboard annotations based on industry standards for film and television creation. Step S7: Structure and store all users' operation trajectories, storyboard annotations, and modification records, generate a unique identifier for each modification, support multi-dimensional query of modification traces, and realize full-process traceability and restoration of modified content; Step S8: Encrypt all collaborative creation data in the cloud and perform multi-node off-site backups.

2. The collaborative management method for an AI video creation platform based on deep learning according to claim 1, characterized in that: In step S1, the deep learning AI model is a user behavior recognition model, which adopts a CNN+LSTM architecture and has been pre-trained with a large amount of video creation user behavior data; the analyzed user historical creation behavior includes editing frequency, annotation type and management operation; the assigned initial collaborative permissions include editing rights, annotation rights, viewing rights and management rights.

3. The collaborative management method for an AI video creation platform based on deep learning according to claim 1, characterized in that: In step S2, visual features of the uploaded scripts, images, and video footage are extracted based on CNN, and the footage is structurally categorized according to shot type, shooting angle, and content. The storyboard list supports drag-and-drop sorting of shots, categorization by scene, and keyword filtering.

4. The collaborative management method for an AI video creation platform based on deep learning according to claim 1, characterized in that: In step S3, the WebSocket real-time communication protocol is used to realize real-time data transmission of the cloud collaborative creation canvas; the real-time data incremental update algorithm only transmits the modified partial data instead of the full data; the multi-person collaborative synchronization delay of the cloud collaborative creation canvas is less than 0.5 seconds.

5. The collaborative management method for an AI video creation platform based on deep learning according to claim 1, characterized in that: In step S4, the storyboard annotations are displayed in red next to the storyboard screen, and the completion status is displayed in green. All authorized users can view the storyboard modification progress, modification content, and annotation comments in real time.

6. The collaborative management method for an AI video creation platform based on deep learning according to claim 1, characterized in that: In step S5, the shared link supports configuring editing or viewing rights, and can be shared across multiple terminal devices such as computers, tablets, and mobile phones; Cross-platform editing data synchronization latency is less than 1 second.

7. The collaborative management method for an AI video creation platform based on deep learning according to claim 1, characterized in that: In step S6, the intelligent reminder methods include platform messages and SMS reminders; the intelligent optimization of storyboard annotation modification opinions is to transform vague modification opinions into specific modification suggestions that conform to the standards of the film and television creation industry.

8. The collaborative management method for an AI video creation platform based on deep learning according to claim 1, characterized in that: In step S7, a unique identifier is generated for each modification as a time-user identifier; the stored modification information includes the modification content, modification time, operator, and state before and after modification; it supports querying modification traces by time, user, and storyboard number.

9. The collaborative management method for an AI video creation platform based on deep learning according to claim 1, characterized in that: In step S8, the collaborative creation data is encrypted and stored in the cloud using the AES256 encryption algorithm; multi-node off-site backups are performed on Alibaba Cloud, Tencent Cloud, and Huawei Cloud, with incremental backups performed every 5 minutes.

10. The collaborative management method for an AI video creation platform based on deep learning according to claim 1, characterized in that: Integrated into the Xcine film and television-grade AI creation platform, it is designed for professional film and television production teams, enabling multi-person collaborative creation, storyboard annotation and modification, and cross-platform editing for directors, cinematographers, storyboard artists, production staff, editors, and production managers; the operation of importing selected shots into the storyboard list takes less than 1 second per shot.

Images