Video Block GPM Index Derivation for Lower Bitrate Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video coding techniques face challenges in improving coding efficiency and quality, particularly in handling geometric partition modes and inter predictions, and extending the application scenarios of template-matching-based methods.
Innovation Solution
The proposed methods derive geometric partition mode (GPM) partition indices and mode information at the decoder or encoder side without indicating them in the bitstream, utilize hybrid prediction tools, and allow templates outside picture boundaries to enhance coding efficiency and quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GPM partition index is signaled in the bitstream, then the decoder can accurately reconstruct the geometric partition mode, but the coding bitrate increases
Solution Approach 1:
The decoder independently derives the GPM partition index using the same geometric partitioning rules and candidate list construction methods as the encoder, eliminating the need for explicit signaling of this index in the bitstream
Solution Approach 2:
The decoder replicates the encoder's GPM partition derivation process by constructing identical candidate lists and applying the same selection criteria, ensuring both sides arrive at the same partition configuration without transmitting the index
2Manufacturing precision
If hybrid prediction tools are implemented, then coding quality improves through blending intra and inter predictions, but device complexity increases
Solution Approach 1:
The system dynamically selects between intra prediction, inter prediction, and their combinations based on the prediction mode flag and candidate availability, adapting the prediction strategy to each specific video block rather than using a fixed complex method for all blocks
Solution Approach 2:
The hybrid prediction mechanism changes the prediction parameter (mode selection) based on conditions such as whether inter prediction is available and the value of the prediction mode flag, switching between different prediction strategies to optimize quality while managing complexity
3Adaptability or versatility
If template matching extends outside picture boundary, then the application scenario and coding efficiency improve, but handling complexity increases
Solution Approach 1:
The system performs preliminary checks to determine whether template matching candidates extend outside the picture boundary before proceeding with the matching process, allowing early termination or alternative handling for out-of-bound cases
Solution Approach 2:
The system applies different handling strategies for different regions: templates completely within the picture boundary use standard matching, while templates extending outside trigger boundary-aware processing or candidate rejection, treating each case according to its specific spatial characteristics
Data Source
AI summary
Embodiments of the present disclosure provide a method for video processing. The method comprises: determining, during a conversion between a current video block of a video and a bitstream of the video, a geometric partition mode (GPM) partition index for the current video block, the GPM partition index being absent from the bitstream; and performing the conversion based on the GPM partition index. Compared with conventional solutions, the proposed method can advantageously reduce the coding bits while maintaining the coding quality.