History-Based Candidate Selection for Reduced-Load Picture Decoding

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing picture coding technologies, such as those using affine transforms for inter prediction, face high processing loads due to complex computations.

Innovation Solution

A picture decoding device and method that utilizes spatial and history-based candidate derivation units to manage candidate lists efficiently, allowing for inter prediction with reduced processing load by selectively adding history-based candidates based on prediction modes.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If affine transform is applied for inter prediction to handle object deformation, then coding efficiency is improved, but processing load becomes great

Engineering Contradiction:

Improvecoding efficiencyVSAvoidprocessing load

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent extracts the complex affine transform computation and replaces it with a simplified candidate list-based approach. Instead of performing full affine transforms, the system derives and selects from pre-computed candidate motion vectors, thereby maintaining coding efficiency while significantly reducing processing load.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent performs preliminary derivation of candidate motion vectors from neighboring blocks and historical data before the actual prediction step. By preparing candidate lists in advance through spatial and temporal neighboring block analysis, the system avoids complex real-time computations while maintaining prediction accuracy.

Inventive Principle:

Principle #10Preliminary action

2Measurement precision

If history-based candidate is added to candidate list, then prediction accuracy is improved, but candidate list management complexity increases

Engineering Contradiction:

Improveprediction accuracyVSAvoidcandidate list management

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent applies different candidate addition rules based on the specific prediction mode (merge mode vs. motion vector predictor mode). This localized approach allows history-based candidates to be selectively added only when beneficial, maintaining prediction accuracy while avoiding unnecessary complexity in candidate list management for all modes.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The patent dynamically adjusts candidate list management behavior based on prediction mode requirements. The system flexibly controls whether history-based candidates are added, removed, or retained depending on the current prediction context, optimizing the balance between accuracy and management complexity for each specific scenario.

Inventive Principle:

Principle #15Dynamics

Data Source

PatentUS12425635B2Picture decoding device, picture decoding method, and picture decoding program with history-based candidate selection

Publication Date: 2025.09.23 GODO KAISHA IP BRIDGE 1

AI summary

A picture decoding device includes a spatial candidate derivation unit configured to derive a spatial candidate from inter prediction information of a block neighboring a decoding target block and register the derived spatial candidate as a candidate to a first candidate list, a history-based candidate derivation unit configured to generate a second candidate list by adding a history-based candidate included in a history-based candidate list as a candidate to the first candidate list, a candidate selection unit configured to select a selection candidate from candidates included in the second candidate list, and an inter prediction unit configured to perform inter prediction using the selection candidate.