Affine Motion Prediction Refinement Through Optical-Flow Block Sizing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding standards face challenges in efficiently managing deblocking filters and quantization parameters, leading to suboptimal video encoding and decoding processes, particularly in the context of emerging standards like Versatile Video Coding (VVC).
Innovation Solution
The proposed methods involve refining deblocking filter application and quantization parameter handling through syntax element-based rules, including conditional overriding and explicit signaling, to enhance video encoding and decoding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If deblocking filter control is added at slice level in addition to picture parameter set level, then flexibility and adaptability of deblocking filter application is improved, but bitstream complexity and parsing overhead increase
Solution Approach 1:
The patent applies segmentation by dividing the deblocking filter control into multiple hierarchical levels: picture parameter set level (pps_deblocking_filter_disabled_flag) and slice level (slice_deblockingFilterDisabled_flag). This allows independent control of deblocking filter at different granularities, enabling flexible adaptation while maintaining structured organization to manage complexity
Solution Approach 2:
The patent implements dynamic control by allowing the deblocking filter status to be overridden at the slice level based on the picture parameter set level setting. The logic dynamically determines whether to apply deblocking filter based on both hierarchical levels, enabling adaptive behavior that responds to different coding conditions
2Measurement precision
If multiple syntax elements are used to control deblocking filter at different levels, then precision of control is improved, but decoding complexity increases
Solution Approach 1:
The patent applies preliminary action by establishing the picture parameter set level deblocking filter control before processing individual slices. This pre-configures the default behavior, and slice-level syntax elements then only need to override when necessary, simplifying the decoding process through hierarchical initialization
Solution Approach 2:
The patent implements feedback through the interaction between picture parameter set level and slice level syntax elements. The slice level control references and overrides the picture parameter set level setting, creating a feedback mechanism that ensures consistent and precise control while managing complexity through structured interaction
3Adaptability or versatility
If deblocking filter override capability is added at slice level, then adaptability of video coding is improved, but processing time increases
Solution Approach 1:
The patent segments the video picture into slices that can independently control deblocking filter application. This segmentation enables parallel processing of different slices and allows the decoder to quickly determine deblocking status based on slice headers, reducing overall processing time while maintaining adaptability
Solution Approach 2:
The patent applies partial action by implementing deblocking filter override capability only where needed at the slice level, rather than requiring full reconfiguration at every level. This selective approach maintains adaptability for specific coding conditions while minimizing the processing overhead of implementing the complete override mechanism
Data Source
AI summary
A method includes determining, for a conversion between a video block of a video and a bitstream of the video, a size of prediction block corresponding to the video block according to a rule. The method also includes performing the conversion based on the determining. The rule specifies that a first size of the prediction block is determined responsive to whether a prediction refinement using optical flow technique is used for coding the video block. The video block has a second size and is coded using an affine merge mode or an affine advanced motion vector prediction mode.