History-Based Motion Vector Prediction With Default Candidates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding standards face challenges in efficiently managing motion vector prediction, particularly in high-efficiency video coding (HEVC) and emerging standards like Versatile Video Coding (VVC), due to complexities in merge mode derivation and motion vector difference coding, which impact bandwidth and computational efficiency.
Innovation Solution
The implementation of history-based motion vector prediction (HMVP) with default parameters, including adaptive addition and resetting of motion candidates in the HMVP table, enhances motion vector prediction by utilizing previously coded motion information to optimize merge candidate lists and reduce redundancy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If history-based motion vector prediction (HMVP) table is reset with default parameters, then coding efficiency is improved and computational complexity is reduced, but motion vector prediction accuracy may be compromised
Solution Approach 1:
The HMVP table is pre-filled with default motion vector candidates before actual motion vector prediction begins. These default candidates are derived from co-located blocks in reference pictures, providing initial prediction values that reduce computational complexity while maintaining reasonable prediction accuracy for the start of encoding/decoding processes
2Measurement precision
If HMVP table size is increased to store more motion candidates, then motion vector prediction accuracy is improved, but memory usage and processing overhead increase
Solution Approach 1:
Different regions or time periods in the video encoding/decoding process use different HMVP table configurations. The table size and candidate selection are adapted locally based on the specific coding unit, picture type, and motion characteristics, rather than using a uniformly large table throughout, thus optimizing the balance between accuracy and complexity for each local context
3Adaptability or versatility
If adaptive addition of K HMVP motion candidates is performed, then motion vector prediction adaptability is improved, but processing complexity increases
Solution Approach 1:
The HMVP table configuration is made dynamic rather than static. The number of candidates K added to the HMVP table is adaptively adjusted based on runtime conditions such as motion complexity, picture type (I, P, B frames), and available computational resources. This allows the system to optimize between accuracy and complexity dynamically during video processing
Data Source
AI summary
Video coding and decoding techniques, including history-based motion vector prediction with default parameters, are described. One example video processing method includes resetting, for a conversion between a video unit of a video and a bitstream of the video, a history-based motion vector prediction (HMVP) table of size N by adding K HMVP motion candidates to the HMVP table, where the K HMVP motion candidates have default values, where N and K are positive integers, and where K≤N; and performing the conversion between the video unit and the bitstream using the HMVP table resulting from the resetting.