GPM Blending Region Signaling for Content-Adaptive Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding techniques, such as MPEG-2, MPEG-4, AVC, HEVC, and VVC, face challenges in improving coding efficiency, particularly in adapting the signaling of metrics for blending regions based on video content.
Innovation Solution
Implementing a geometric partitioning mode (GPM) for video units, where the signaling of a metric for blending regions is dependent on the content of the video unit, and sample values in blending regions are determined based on multiple predictions, enabling adaptive signaling and improved coding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a fixed signaling approach is used for blending region metrics in GPM, then the coding structure is simple, but the coding efficiency cannot be adaptively improved based on video content
Solution Approach 1:
The patent applies dynamics by making the signaling of blending region metrics adaptive rather than fixed. The encoder dynamically decides whether to signal the metric based on content characteristics, and the metric value itself is dynamically determined based on the content of the current video unit. This allows the system to adapt its complexity to the actual content requirements, improving coding efficiency when needed while maintaining simplicity for uniform content.
Solution Approach 2:
The patent changes the parameter of the metric value from a fixed predetermined value to a content-dependent variable. The metric value is determined based on the actual content of the current video unit, allowing the system to optimize blending region processing parameters according to the specific content characteristics, thereby improving coding efficiency without requiring complex additional signaling structures.
2Measurement precision
If a predetermined metric value is used for blending regions, then the processing is simple and fast, but the accuracy and adaptability to different video contents are limited
Solution Approach 1:
The patent applies local quality by determining the metric value based on the specific content of each current video unit rather than using a global predetermined value. This allows different regions and content types to have appropriately optimized metric values, improving measurement precision and adaptability to local content characteristics without requiring overly complex determination mechanisms.
Solution Approach 2:
The system determines the metric value based on the content of the current video unit itself, making the system self-adapting. The video content provides the information needed to determine its own optimal metric value, eliminating the need for external complex control mechanisms while achieving content-adaptive precision.
Data Source
AI summary
Embodiments of the present disclosure provide a solution for video processing. A method for video processing is proposed. The method comprises: performing a conversion between a current video unit of a video and a bitstream of the video, wherein the current video unit is coded with a geometric partitioning mode (GPM) based prediction mode, whether a value for a metric of a blending region of the current video unit in a direction is indicated in the bitstream is dependent on content of the current video unit, and values for samples in the blending region are determined based on values for samples in a plurality of predictions associated with the current video unit.