Affine Bi-Prediction Motion Refinement for Video Block Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in achieving efficient compression and reconstruction of video data, particularly in intra prediction and motion compensation, due to the increasing complexity and diversity of prediction directions and motion vectors, which results in suboptimal bit usage and reconstruction quality.
Innovation Solution
The implementation of affine bi-prediction mode with bilateral matching motion refinement, where control point motion vectors and affine parameters are refined using offset values to minimize bilateral matching cost, improving the accuracy of motion compensation and intra prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If traditional motion compensation techniques are used, then device complexity is reduced, but manufacturing precision (reconstruction quality) deteriorates
Solution Approach 1:
The patent refines motion vectors by adjusting their parameter values (offset values) to improve prediction accuracy. The control point motion vectors are modified using refinement offset values to minimize bilateral matching cost, thereby enhancing reconstruction quality without requiring a complete redesign of the motion compensation system.
Solution Approach 2:
The patent performs preliminary refinement of control point motion vectors before final motion compensation. By pre-adjusting the motion vectors using offset values and selecting the best refinement mode in advance, the system optimizes reconstruction quality while avoiding more complex real-time calculations during decoding.
2Manufacturing precision
If more prediction directions and motion vectors are used, then manufacturing precision (prediction accuracy) is improved, but loss of information (bit usage efficiency) worsens
Solution Approach 1:
The patent extracts only the essential refinement information (offset values) from the full motion vector data. Instead of transmitting complete refined motion vectors, the system transmits only the differences (offsets) from the original control point motion vectors, significantly reducing bit usage while maintaining prediction accuracy.
Solution Approach 2:
The patent discards redundant motion vector information and recovers only the necessary refinement data. By calculating bilateral matching costs and selecting only the optimal refinement mode and offset values, the system eliminates unnecessary data transmission while preserving the most important prediction accuracy improvements.
3Manufacturing precision
If affine bi-prediction mode with bilateral matching refinement is implemented, then manufacturing precision (motion compensation accuracy) is improved, but device complexity increases
Solution Approach 1:
The patent segments the motion compensation process into distinct stages: initial control point motion vector derivation, refinement offset calculation, bilateral matching cost computation, and final vector selection. This segmentation allows each stage to be optimized independently, improving motion compensation accuracy while managing device complexity through modular processing.
Solution Approach 2:
The patent applies partial refinement by selecting only the most beneficial refinement mode (first or second mode) based on bilateral matching costs. Instead of applying all possible refinement techniques to every block, the system selectively applies refinement only where it provides the greatest accuracy improvement, balancing device complexity with motion compensation accuracy.
Data Source
AI summary
Aspects of the disclosure provide methods and apparatuses for video encoding/decoding. In some examples, an apparatus for video decoding includes processing circuitry. The processing circuitry receives a coded video bitstream including a current picture. The current picture includes a current block. The processing circuitry determines, based on decoded information from the coded video bitstream, that the current block in the current picture is coded in an affine bi-prediction mode. The current block is predicted by a plurality of control point motion vectors. The processing circuitry refines the plurality of control point motion vectors by applying bilateral matching motion refinement, and reconstructs the current block according to the refined plurality of control point motion vectors associated with the current block.