Flexible TPM Candidate List Signaling for Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding techniques face challenges in efficiently reducing redundancy in video data, particularly in inter-picture prediction, where motion vectors (MVs) often require significant bandwidth and storage, and existing prediction mechanisms do not effectively address the complexity of triangular prediction modes in video compression.
Innovation Solution
The method involves constructing a triangular prediction mode (TPM) candidate list for a current coding block, using a syntax element to indicate the maximum allowed number of TPM candidates, and determining merge indices from motion vectors of merge candidates to process the block with two motion vectors as predictors for triangular partitions, allowing for efficient coding and decoding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If motion vectors are coded directly without prediction, then coding accuracy is maintained, but bandwidth and storage requirements increase significantly
Solution Approach 1:
The patent uses motion vector prediction by copying motion vector information from neighboring blocks and reference pictures. Instead of coding each motion vector directly, the decoder reconstructs motion vectors by copying and interpolating from previously decoded neighboring blocks, significantly reducing the bitrate required while maintaining adequate coding accuracy.
Solution Approach 2:
The patent introduces motion vector prediction as an intermediary mechanism between direct motion vector coding and the final motion compensation. Prediction modes (spatial, temporal, and combined) act as mediators that generate approximate motion vectors from neighboring information, reducing the precision requirements and bandwidth needed for explicit motion vector transmission.
2Quantity of substance
If triangular prediction mode is applied to improve compression efficiency, then bandwidth reduction is achieved, but decoder complexity increases
Solution Approach 1:
The patent divides the current block into two triangular prediction units (TPUs) with different split directions (first and second diagonal directions). This segmentation allows independent motion vector prediction for each triangle, improving compression efficiency while keeping the complexity manageable through systematic processing of each segment.
Solution Approach 2:
The patent implements flexible triangular prediction mode where the decoder can adaptively select between different prediction configurations based on block characteristics. The maximum number of TPM candidates can be configured, allowing dynamic adjustment between compression efficiency and computational complexity based on content requirements.
3Productivity
If maximum number of TPM candidates is increased to improve prediction accuracy, then coding efficiency improves, but processing complexity and memory requirements increase
Solution Approach 1:
The patent introduces a configurable parameter (maximum number of TPM candidates) that controls the size of the candidate list for triangular prediction mode. By adjusting this parameter, the system can optimize the trade-off between coding efficiency and processing complexity, allowing adaptive tuning based on content characteristics and resource constraints.
Data Source
AI summary
A method of video encoding includes determining a first maximum allowed number of triangular prediction mode (TPM) candidates of a TPM applied to a first set of coding blocks, and constructing a TPM candidate list of a current coding block within the first set of coding blocks processed with the TPM according to the determined first maximum allowed number of TPM candidates. The method also includes generating a coded bitstream including the current coding block encoded according to the constructed TPM candidate list and including a first syntax element indicating the first maximum allowed number of TPM candidates.