Picture Decoding Candidate Lists for Lower Motion Prediction Load
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing picture coding technologies, such as those described in Patent Literature 1, suffer from high processing loads due to redundant motion vector prediction checks, which hinder efficient coding and decoding processes.
Innovation Solution
The proposed solution involves a picture decoding device and method that removes redundant checks for motion vector prediction modes in merge modes, allowing for a larger candidate list in merge modes and updating history-based motion vector predictor candidate lists to avoid overlaps, thereby reducing processing load.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If redundant motion vector prediction checks are performed for both merge mode and motion vector predictor mode, then prediction accuracy is improved, but processing load increases significantly
Solution Approach 1:
The patent applies different redundancy check strategies for different prediction modes. For merge mode, redundancy checks are removed to reduce processing load, while for motion vector predictor mode, redundancy checks are maintained to ensure prediction accuracy. This localized differentiation resolves the contradiction by optimizing each mode according to its specific requirements.
Solution Approach 2:
The patent segments the candidate list management into mode-specific handling. The history-based candidate derivation unit operates differently based on the prediction mode, with merge mode using one set of rules and motion vector predictor mode using another set of rules. This segmentation allows independent optimization of each mode's processing characteristics.
2Device complexity
If the same candidate list management rules are applied to both merge mode and motion vector predictor mode, then implementation simplicity is maintained, but processing efficiency decreases due to unnecessary checks
Solution Approach 1:
The patent introduces mode-specific quality attributes in candidate list management. The history-based candidate derivation unit applies different behaviors based on the prediction mode flag, making the implementation simpler for merge mode by removing redundant checks while maintaining thoroughness for motion vector predictor mode.
Solution Approach 2:
The patent makes the candidate list management dynamic by adjusting the derivation rules based on the prediction mode. The system transitions from a static, uniform approach to a dynamic, adaptive approach where the history-based candidate derivation unit modifies its behavior according to the current prediction mode, optimizing processing efficiency for each mode.
3Speed
If history-based candidates are added to the candidate list without checking for overlaps, then processing speed increases, but prediction quality may deteriorate due to duplicate candidates
Solution Approach 1:
The patent applies selective overlap checking based on prediction mode. For merge mode, overlap checking is removed to maximize processing speed, while for motion vector predictor mode, overlap checking is maintained to ensure prediction quality. This localized approach resolves the contradiction by optimizing each mode according to its specific performance requirements.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Technology for improving coding efficiency by performing a block split suitable for picture coding and decoding is provided. 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 in 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. The history-based candidate derivation unit switches between whether or not a history-based candidate overlapping a candidate included in the first candidate list is added in accordance with a prediction mode.