Personalized Object Model Adapters for Video Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video compression schemes are inefficient, leading to slow data communication speeds, large storage requirements, and perceptual disturbances, particularly in handling video data that is sensitive to visual information.
Innovation Solution
The system processes video signals to create object models that can be archived and used with a codec to reconstruct compressed video files, allowing for efficient compression, storage, and transmission by identifying and reducing redundant models, and enabling customized codecs based on user-specific data for improved quality and access control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional video compression schemes are used, then video data can be stored and transmitted, but data communication speeds are slow and storage requirements are large
Solution Approach 1:
The video signal is segmented into multiple objects, and object models are created for each object. This segmentation allows the system to process and compress only the essential characteristics of each object rather than the entire video frame, reducing storage requirements and improving communication speed.
Solution Approach 2:
Object models are created as compressed representations (copies) of the actual video objects. These object models serve as simplified copies that capture the essential features of video objects, allowing efficient storage and transmission while maintaining the ability to reconstruct the original video content.
2Loss of information
If conventional video compression schemes are used, then video data can be compressed, but perceptual disturbances occur particularly in visually sensitive information
Solution Approach 1:
The system applies different processing and compression strategies to different objects in the video based on their visual importance. Objects that are more visually sensitive receive higher quality processing and less compression, while less critical objects can be compressed more aggressively. This local quality approach maintains overall video quality while reducing perceptual disturbances.
3Productivity
If object models are archived and customized codecs are created, then video processing efficiency is improved, but system complexity increases
Solution Approach 1:
Object models are created and archived in advance during the encoding phase. This preliminary action allows the system to have pre-processed object models available for efficient video reconstruction during playback, improving processing efficiency while the complexity is managed during the initial model creation phase.
Solution Approach 2:
The system creates customized codecs automatically by grouping archived object models based on user-specific data and video content analysis. This self-service approach allows the system to adapt to different users and video types without requiring manual codec configuration, improving efficiency while managing complexity through automation.
Data Source
AI summary
Personal object based archival systems and methods are provided for processing and compressing video. By analyzing features unique to a user, such as face, family, and pet attributes associated with the user, an invariant model can be determined to create object model adapters personal to each user. These personalized video object models can be created using geometric and appearance modeling techniques, and they can be stored in an object model library. The object models can be reused for processing other video streams. The object models can be shared in a peer-to-peer network among many users, or the object models can be stored in an object model library on a server. When the compressed (encoded) video is reconstructed, the video object models can be accessed and used to produce quality video with nearly lossless compression.