Affine-Translational Merge Prediction for Complex Video Motion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in efficiently utilizing intra prediction and motion compensation techniques to reduce redundancy in video data, particularly in handling complex motion vectors and affine transformations, which impact compression efficiency and quality.
Innovation Solution
The implementation of affine-translational merge candidates in video encoding/decoding, which combine affine motion information from one reference picture with translational motion information from another, allowing for more accurate prediction and reduced bit requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If affine motion compensation with multiple reference pictures is used, then prediction accuracy is improved, but bit requirements increase
Solution Approach 1:
The motion information is segmented into two distinct types: affine motion information from a first reference picture and translational motion information from a second reference picture. This segmentation allows the decoder to process different motion models separately, reducing the complexity of transmitting and processing combined motion parameters while maintaining high prediction accuracy through the use of multiple reference pictures with different motion characteristics
Solution Approach 2:
The patent introduces an additional dimension by combining two different motion models (affine and translational) from two different reference pictures. Instead of using a single complex motion model, the solution adds a temporal dimension by merging motion information across multiple reference pictures, achieving higher prediction accuracy without proportionally increasing bit requirements
2Measurement precision
If complex motion vectors are encoded, then motion precision is improved, but compression efficiency deteriorates
Solution Approach 1:
The patent applies partial action by selecting only the necessary components of motion information from multiple reference pictures. Instead of encoding complete motion vectors for all reference pictures, the method selectively uses affine motion information from one reference picture and translational motion information from another, achieving sufficient motion precision while reducing the amount of data that needs to be encoded and transmitted
3Measurement precision
If multiple reference pictures are utilized, then prediction quality is improved, but processing complexity increases
Solution Approach 1:
The patent merges motion information from multiple reference pictures by combining affine motion parameters from a first reference picture with translational motion parameters from a second reference picture. This merging approach allows the system to leverage the strengths of different motion models without requiring separate processing pipelines, thereby improving prediction quality while managing processing complexity through unified motion compensation logic
Data Source
AI summary
In some examples, an apparatus for video decoding includes processing circuitry. The processing circuitry determines, from a candidate list, a first affine-translational merge candidate for a prediction of a current block in a current picture. The first affine-translational merge candidate provides affine motion information associated with a first reference picture in a first reference list and translational motion information associated with a second reference picture in a second reference list. The processing circuitry generates a first prediction for a sample in the current block according to the affine motion information associated with the first reference picture, and generates a second prediction for the sample in the current block according to the translational motion information associated with the second reference picture. The processing circuitry reconstructs the sample of the current block according to a combination of the first prediction and the second prediction.