HMVP Weight Index Inheritance for Bi-Directional Motion Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face inefficiencies in predicting motion vectors, particularly in bi-directional prediction scenarios, leading to suboptimal compression ratios and increased data requirements.
Innovation Solution
Implementing a History-based Motion Vector Prediction (HMVP) method that stores motion information candidates with weight parameters for bi-directional and uni-directional predictions, allowing for dynamic weight adjustments based on block relationships and reference picture configurations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If motion vector prediction is performed using conventional methods, then data requirements are reduced, but compression ratio and prediction accuracy deteriorate in bi-directional scenarios
Solution Approach 1:
The patent pre-calculates and stores weight parameters for motion vector prediction in a history list during the encoding of previous blocks. These pre-computed weight parameters are then directly reused for current block prediction, eliminating the need for complex real-time calculations and improving both compression ratio and prediction accuracy while managing complexity through temporal reuse
Solution Approach 2:
The patent implements a feedback mechanism where motion vector prediction results and corresponding weight parameters from previously encoded blocks are stored in a history list. This historical information is continuously fed back and reused for subsequent block predictions, creating a self-improving system that adapts to local motion patterns and enhances compression efficiency
2Measurement precision
If weight parameters are stored for each motion information candidate, then prediction accuracy improves, but memory requirements and data structure complexity increase
Solution Approach 1:
The patent applies local quality by associating specific weight parameters with individual motion information candidates in the history list. Each candidate maintains its own optimized weight parameter that reflects local motion characteristics, enabling precise prediction while avoiding the need to store uniform data structures for all candidates. This localized parameter storage optimizes the balance between accuracy and memory usage
3Productivity
If conventional motion vector prediction is used, then encoding complexity is reduced, but compression efficiency and bitrate performance worsen
Solution Approach 1:
The patent introduces weight parameters as additional degrees of freedom in the motion vector prediction process. By modifying the prediction model to include these parameters and storing them in the history list, the system achieves superior compression efficiency. The increased encoding complexity is managed through systematic parameter management and temporal reuse, making the added complexity worthwhile for the significant bitrate savings achieved
Data Source
AI summary
Aspects of the disclosure provide methods and apparatuses for video coding. In some examples, an apparatus includes processing circuitry. The processing circuitry obtains prediction information of a first block in a picture from a coded video bitstream, and generates reconstructed samples of the first block according to the prediction information and one of bi-directional prediction and uni-directional prediction. The processing circuitry adds motion information and a bi-prediction weight index of a History-based Motion Vector Prediction (HMVP) candidate to an HMVP list based on the prediction information of the first block and whether the first block is coded according to the bi-directional prediction or the uni-directional prediction. Further, the processing circuitry generates reconstructed samples of a second block in the picture based on a plurality of candidates that includes the HMVP candidate.