Inter-Prediction Merge Refinement for High-Resolution Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video encoding technologies face challenges in efficiently processing next-generation video contents with high spatial resolution, high frame rate, and high dimensionality of scene representation, leading to increased memory storage and processing demands.
Innovation Solution
A method for generating refined merge candidates using motion vector offsets and weighted averaging, enhancing the merge candidate list for improved prediction accuracy and compression performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional video encoding methods are used for next-generation video contents, then basic compression is achieved, but prediction accuracy and compression performance are insufficient for high spatial resolution and high frame rate contents
Solution Approach 1:
The patent changes the parameters of merge candidates by generating refined merge candidates through motion vector offset addition and weighted averaging operations. This transforms the original merge candidate parameters (motion vectors) into refined parameters that provide better prediction accuracy for high-resolution video content, directly addressing the insufficiency of conventional compression methods.
Solution Approach 2:
The patent segments the merge candidate generation process into multiple stages: configuring initial merge candidates from spatial and temporal neighbors, generating refined candidates through offset addition, and creating additional refined candidates through weighted averaging. This segmentation allows systematic improvement of prediction accuracy while managing computational complexity.
2Measurement precision
If the merge candidate list is enhanced with refined candidates, then prediction accuracy improves, but encoding complexity increases
Solution Approach 1:
The patent performs preliminary actions by pre-configuring merge candidate lists with refined candidates before the actual encoding process. The refined merge candidates are generated in advance using motion vector offsets and weighted averaging, so that during encoding, the decoder can directly use these pre-computed candidates without performing complex calculations, thus reducing real-time encoding complexity while maintaining high prediction accuracy.
3Productivity
If motion vector refinement is applied to improve prediction accuracy, then compression performance increases, but processing power requirements increase
Solution Approach 1:
The patent introduces motion vector offsets as intermediaries to refine motion vectors. Instead of performing complex motion estimation for each merge candidate, the system uses offset values (which can be derived from neighboring blocks or temporal information) to adjust base motion vectors. This intermediary approach significantly reduces processing power requirements while achieving the compression performance needed for high-resolution and high-frame-rate video content.
Data Source
AI summary
Disclosed are a method for decoding a video signal and an apparatus therefor. Specifically, a method for decoding an image based on an inter-prediction mode may include: configuring a merge candidate list by using motion information of a spatial candidate and a temporal candidate adjacent to a current block; parsing a first syntax element indicating a particular candidate applied to inter-prediction of the current block among a predefined number of candidates in the merge candidate list; deriving a refined candidate by adding a motion vector offset to a motion vector of a candidate indicated by the first syntax element; and generating a prediction block of the current block by using the motion information of the refined candidate.