Video Coding Block Size Dependent Inter Prediction Bandwidth Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding standards face challenges in reducing bandwidth and line buffers, particularly with the introduction of new coding tools like affine prediction and bi-prediction, which increase memory bandwidth requirements and line buffer usage.
Innovation Solution
The implementation of specific techniques such as constraining motion vectors within pre-defined fields, clipping motion vectors to reduce bandwidth, and modifying block sizes and interpolation methods to optimize memory usage, along with disabling certain coding tools based on block dimensions, to manage bandwidth and line buffers effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If affine prediction and bi-prediction coding tools are introduced to improve coding performance, then video quality and compression efficiency are improved, but memory bandwidth requirements and line buffer usage increase
Solution Approach 1:
The current block is divided into multiple sub-blocks, and motion vectors are constrained within pre-defined fields for each sub-block. This segmentation allows independent processing of motion compensation for each sub-block, reducing the overall memory bandwidth requirement while maintaining coding efficiency through localized motion prediction
Solution Approach 2:
Different motion vector constraints and prediction modes are applied to different regions (sub-blocks) of the current block based on their local characteristics. This local quality approach optimizes memory usage by applying appropriate prediction complexity to each region, reducing overall memory bandwidth while preserving important local motion details
2Measurement precision
If new coding tools like affine prediction are implemented to improve prediction accuracy, then coding performance is improved, but line buffer requirements increase
Solution Approach 1:
The block is segmented into sub-blocks with independent motion vector constraints, allowing affine prediction to be applied locally where needed. This reduces the line buffer requirement by processing smaller regions independently rather than requiring large buffers for entire blocks
Solution Approach 2:
Affine prediction and bi-prediction are applied selectively to specific sub-blocks rather than uniformly to the entire block. This partial application maintains prediction accuracy for regions that benefit most while reducing the overall line buffer requirements
Data Source
AI summary
Techniques for implementing video processing techniques are described. In one example implementation, a method of video processing includes determining, for a conversion between a current block of a video and a bitstream representation of the video, a manner in which coding information of the current block is represented in the bitstream representation in part based on whether a condition associated with a size of the current block is satisfied. The method also includes performing the conversion based on the determining.