On-Demand Avatar Video Rendering for Personalized Media Playlists
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems face challenges in generating personalized video content featuring user avatars on demand, especially with millions of users, due to the need for real-time customization and up-to-date avatars, which is difficult to achieve with pre-generated content.
Innovation Solution
The system renders video content including user avatars on demand by replacing placeholder avatars with personalized user avatars in real-time, using a content controller system that includes a delivery server and rendering server to manage and update avatar incorporation into media content segments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple avatars are generated and stored for different scenarios, then the variety and quality of video content is improved, but the storage space requirements and system complexity increase
Solution Approach 1:
The system segments avatar generation by separating the base avatar model from scenario-specific variations. Instead of storing complete distinct avatars for each scenario, the system stores a base avatar and generates scenario-specific variations on-demand through controlled transformation, reducing storage requirements while maintaining content variety.
Solution Approach 2:
The system creates virtual copies of the base avatar by applying scenario-specific transformations rather than storing completely different avatars. This copying approach allows multiple scenario variations to be generated from a single source, reducing the need for extensive storage of unique avatar data.
2Productivity
If avatars are pre-generated and stored for various scenarios, then video generation speed is improved, but the system complexity and resource consumption increase
Solution Approach 1:
The system performs preliminary generation of a base avatar and pre-computes transformation parameters for different scenarios. During actual video generation, the system applies pre-defined transformations to the base avatar rather than generating complete avatars from scratch, reducing real-time computational complexity while maintaining generation speed.
Solution Approach 2:
The system transitions from static pre-generated avatars to dynamic on-demand generation using a base avatar combined with scenario-specific transformations. This dynamic approach allows the system to adapt to different scenarios flexibly without requiring complex pre-computed assets for each possible scenario.
3Device complexity
If a single base avatar is used for all scenarios, then storage space and system complexity are reduced, but the diversity and quality of video content deteriorate
Solution Approach 1:
The system applies local quality transformations to the base avatar by adjusting specific parameters (pose, expression, lighting, background) according to scenario requirements. This allows different regions or aspects of the avatar to be optimized for specific scenarios while maintaining the overall structure of a single base avatar, achieving diversity without proportionally increasing complexity.
4Volume of stationary object
If avatars are generated on-demand from a base model, then storage space is reduced, but video generation time and computational resources increase
Solution Approach 1:
The system performs partial generation by applying only the necessary transformations to the base avatar for each specific scenario rather than regenerating the complete avatar from scratch. This partial action approach reduces computational time and resources while maintaining the quality and diversity of the generated content.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Method starts with a processor causing an electronic program guide (EPG) to be displayed on a first client device that is associated with a first user. Processor detects a selection of a selectable item included in the EPG associated with a media content item by the first user. Processor identifies a second user that is connected to the first user on a communication platform. Processor retrieves a main user identification associated with the first user and a secondary user identification associated with the second user and generates a request for the media content item including media content identification, main user identification, and secondary user identification. Processor then receives a playlist associated with the media content item that comprises media content item segments including first avatar associated with first user and second avatar associated with second user. Other embodiments are disclosed herein.