Local Illumination Compensation in UMVE Inter Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding technologies face challenges in achieving better compression ratios and lower complexity, particularly in handling illumination changes and motion vector prediction, which affect encoding efficiency and decoding performance.
Innovation Solution
Incorporating local illumination compensation (LIC) and advanced motion vector prediction tools, such as multi-hypothesis inter prediction and weighted prediction, to enhance video encoding and decoding processes, utilizing linear models for illumination changes and adaptive methods for motion vector refinement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If local illumination compensation (LIC) tool is integrated with multiple prediction tools (IBC, GBI, multi-hypothesis prediction), then video coding efficiency and compression ratio are improved, but device complexity and processing complexity increase
Solution Approach 1:
The patent combines the LIC tool with multiple prediction tools (IBC, GBI, multi-hypothesis prediction) into a unified video coding framework. The LIC module integrates with these prediction tools to jointly process illumination compensation and prediction, allowing the system to leverage multiple prediction strategies while applying illumination correction across all of them, thereby improving overall coding efficiency despite increased complexity
Solution Approach 2:
The LIC tool is designed as a universal compensation mechanism that can be applied across different prediction modes and tools. Rather than creating separate illumination compensation paths for each prediction tool, the patent implements a multi-functional LIC module that serves all prediction tools (IBC, GBI, multi-hypothesis prediction), reducing redundant complexity while maintaining improved coding efficiency
2Adaptability or versatility
If LIC tool is used with adaptive selection rules (exclusive with CPR mode, inheritance from UMVE mode), then coding flexibility and adaptability are improved, but processing overhead and computational burden increase
Solution Approach 1:
The patent implements dynamic adaptive selection rules for LIC tool usage. The system dynamically determines whether to apply LIC based on the active prediction mode (e.g., exclusive with CPR mode, inheritance from UMVE mode). These adaptive rules allow the system to flexibly enable or disable LIC processing depending on the current coding context, improving adaptability while managing processing overhead through conditional execution rather than constant processing
Data Source
AI summary
A method for video processing is provided to include: determining, based on an inheritance rule, whether to use a local illumination compensation (LIC) mode for a conversion between a current block of a video region of a video and a coded representation of the video, and performing the conversion based on the determining. The LIC mode includes using a linear model of illumination changes in the current block during the conversion. The inheritance rule specifies to inherit a use of the LIC mode from a base merge candidate of an ultimate motion vector expression (UMVE) mode, and the UMVE mode includes using a motion vector expression that includes a base merge candidate and motion vector differences utilized to refine the motion information of the base merge candidate.