Motion Vector Predictor List Ordering for Accurate Video Coding

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

The ordering process of motion vector prediction (MVP) lists in video coding is sub-optimal, as important motion vector candidates are placed in less favorable positions, leading to inefficient video compression.

Innovation Solution

Insert motion vector candidates from a reference MV bank into the MVP list before derived candidates when certain conditions are met, prioritizing spatial MV candidates for improved accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If a pre-defined weighting-based re-ordering process is used to re-order only adjacent spatial MVP, temporal MVP, and non-adjacent SMVP, then the processing is simplified, but the motion vector prediction accuracy deteriorates because other MV candidates are not re-ordered

Engineering Contradiction:

Improvesimplicity of re-ordering processVSAvoidmotion vector prediction accuracy

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

The patent segments the MV candidates into different groups (spatial MVP candidates, temporal MVP candidates, non-adjacent SMVP candidates, and other MV candidates) and applies different re-ordering strategies to each group. This allows the system to maintain simplicity for commonly used candidates while extending re-ordering benefits to all candidates, resolving the contradiction between operational simplicity and prediction accuracy.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces a dynamic re-ordering mechanism that adapts the MVP list based on actual prediction performance. The system continuously monitors which MV candidates provide the best prediction accuracy and dynamically adjusts their positions in the MVP list, allowing the system to optimize accuracy without requiring complex pre-defined re-ordering rules for all candidate types.

Inventive Principle:

Principle #15Dynamics

2Device complexity

If MV candidates from the reference MV candidate bank are inserted at the end of the MVP list, then the list construction is simpler, but the prediction accuracy deteriorates because less accurate candidates are prioritized

Engineering Contradiction:

ImproveMVP list construction complexityVSAvoidmotion vector prediction accuracy

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The patent performs preliminary actions by pre-ordering MV candidates from the reference MV candidate bank based on their potential prediction accuracy before inserting them into the MVP list. This preliminary sorting ensures that the most accurate candidates are positioned earlier in the list, improving prediction accuracy without significantly increasing construction complexity.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent applies local quality by treating different MV candidate groups differently based on their specific characteristics and prediction potential. Spatial MVP candidates receive different handling than temporal MVP candidates, and the reference MV candidate bank is processed with specific prioritization rules that consider local prediction accuracy characteristics of each candidate type.

Inventive Principle:

Principle #3Local quality

3Measurement precision

If spatial MV candidates are prioritized in the MVP list, then prediction accuracy improves, but the system complexity increases due to additional re-ordering operations

Engineering Contradiction:

Improvemotion vector prediction accuracyVSAvoidMVP list construction complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments MV candidates into distinct groups based on their source and characteristics (spatial, temporal, non-adjacent, and other candidates). By segmenting the candidates, the system can apply targeted re-ordering operations to each group according to their specific prediction characteristics, improving accuracy while managing complexity through structured processing.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent changes the ordering parameters of MV candidates in the MVP list based on their prediction accuracy characteristics. By adjusting the position and weighting parameters of different candidate types, the system optimizes prediction accuracy. The parameter changes are applied systematically to maintain balance between accuracy improvement and system complexity.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20250324077A1Systems and methods for motion vector predictor list improvements

Publication Date: 2025.10.16 TENCENT AMERICA LLC

AI summary

The various implementations described herein include methods and systems for encoding and decoding video blocks. An example method includes retrieving one or more motion vector (MV) candidates from a reference MV bank. The method includes, in accordance with a determination that a first condition is satisfied, inserting the one or more MV candidates from the reference MV bank into the MVP list associated with the current coding block after derived MV candidates is inserted into the MVP list. The method includes, in accordance with a determination that the first condition is not satisfied, inserting the one or more MV candidates from the reference MV bank into the MVP list associated with the current coding block before the derived MV candidates is inserted into the MVP list.