Motion Vector Predictor Coding with Adaptive CABAC Contexts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing video coding standards like HEVC face challenges in efficiently encoding and decoding motion vectors due to the complexity and inefficiency of tools such as affine motion mode and Alternative Temporal Motion Vector Prediction (ATMVP), particularly in handling complex motions and increased coding complexity.
Innovation Solution
Implementing methods and devices for encoding and decoding motion vector predictor indices using Context Adaptive Binary Arithmetic Coding (CABAC) with techniques like bypass coding or shared contexts, and context variables based on neighboring blocks or block complexity to optimize the encoding and decoding processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If Alternative Temporal Motion Vector Prediction (ATMVP) is enabled to improve motion vector prediction accuracy, then coding efficiency is improved, but device complexity increases due to increased number of merge candidates
Solution Approach 1:
The patent applies parameter changes by modifying the CABAC coding parameters (context variables) based on the block type and prediction mode. Different context variables are selected for different blocks (e.g., left, top, top-left neighbors) and different prediction modes (inter, affine inter), allowing the system to adapt the coding parameters to the specific characteristics of each block without increasing structural complexity
Solution Approach 2:
The system dynamically adjusts the CABAC coding approach by selecting different context variables based on the block's spatial position and prediction mode. The decoding process dynamically determines which context variable to use based on previously decoded information, enabling adaptive coding efficiency without requiring complex pre-configured structures for every possible block type
2Productivity
If more CABAC bits are coded to improve prediction accuracy, then coding efficiency is improved, but encoding time increases
Solution Approach 1:
The patent applies local quality by treating different bits of the motion vector predictor index differently based on their position and importance. The first bit (most significant) uses CABAC coding with context variables to capture important prediction information, while subsequent bits use bypass CABAC coding for faster encoding. This localized differentiation allows the system to spend encoding resources where they provide the most benefit
Solution Approach 2:
The system applies partial action by selectively applying full CABAC coding only to the first bit of the motion vector predictor index, while using the simpler bypass CABAC coding for the remaining bits. This partial application of the more complex coding method achieves sufficient coding efficiency for the most important prediction information while avoiding the time cost of applying full CABAC to all bits
Data Source
AI summary
A method of encoding information about a motion information predictor, comprising: selecting one of a plurality of motion information predictor candidates; and encoding information for identifying the selected motion information predictor candidate using CABAC coding, wherein the CABAC coding comprises, for at least one bit of said information, using the same context variable used for another inter prediction mode when one or both of a Triangle Merge mode or a Merge with Motion Vector Difference (MMVD) Merge mode is used.