Video Block Prediction Fusion With Adaptive Template Weights
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in improving the accuracy of intra and inter prediction methods to enhance video compression efficiency.
Innovation Solution
The proposed solution involves determining a template of a current block and one or more prediction templates, calculating weights for these templates based on the current block, and fusing the prediction blocks to obtain a second prediction block, thereby enhancing prediction accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional intra prediction and inter prediction methods are used, then video coding can be performed with basic compression, but prediction accuracy is insufficient leading to higher redundancy
Solution Approach 1:
The patent changes the parameters of prediction by introducing multiple prediction templates (first, second, third prediction templates) with different characteristics. Each template is assigned a weight value that can be adjusted based on its prediction accuracy for the current block. This parameter adjustment allows the system to adaptively select the most suitable prediction template, thereby improving prediction accuracy and reducing residual redundancy without requiring complex additional structures.
Solution Approach 2:
The patent creates a composite prediction structure by combining multiple prediction templates into a unified prediction framework. Instead of using a single prediction template, the system integrates multiple templates with different prediction characteristics (intra prediction, inter prediction, and combined prediction templates) and fuses them through weighted combination. This composite approach leverages the strengths of different prediction methods to achieve higher overall prediction accuracy and better compression efficiency.
2Measurement precision
If multiple prediction templates are introduced to improve prediction accuracy, then prediction performance improves, but calculation complexity increases
Solution Approach 1:
The patent manages calculation complexity by introducing a weight parameter system that simplifies the selection and combination process. Instead of evaluating all possible prediction templates equally, the system assigns weight values to different templates based on their predicted performance. This parameter-based approach allows the encoder to focus computational resources on calculating weights for a limited set of candidate templates rather than exhaustively processing all possible prediction modes, thereby balancing accuracy improvement with computational feasibility.
3Measurement precision
If adaptive weight determination is used for prediction templates, then prediction accuracy improves, but encoding complexity increases
Solution Approach 1:
The patent optimizes encoding efficiency by changing the parameter representation of template weights. Instead of encoding full-precision weight values, the system uses quantized weight indices that reference pre-defined weight tables. This parameter transformation reduces the number of bits required to encode weight information while maintaining sufficient prediction accuracy. The decoder can efficiently reconstruct the weight values from these compact indices, thereby improving overall encoding efficiency without significantly compromising prediction performance.
Data Source
AI summary
Disclosed in the embodiments of the present application are an encoding method and apparatus, a decoding method and apparatus, an encoder, a decoder, a code stream, and a storage medium. The encoding method is applied to an encoder, and comprises: determining a template of the current block and one or more prediction templates of the template of the current block; determining the weights of the one or more prediction templates according to the template of the current block and the one or more prediction templates; determining one or more first prediction blocks of the current block according to a prediction parameter of the current block; and fusing the one or more first prediction blocks according to the weights of the one or more prediction templates, so as to obtain a second prediction block of the current block.