Transform Coefficient Sign Prediction With Transform-Specific CABAC Contexts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding techniques face inefficiencies in entropy coding of sign prediction errors due to the lack of adaptation to different transforms and contexts, leading to increased complexity and reduced coding efficiency.
Innovation Solution
Adopting a multiple core transform technique that allows selection of discrete cosine transforms (DCT) and discrete sine transforms (DST) for entropy coding, with context adaptation based on transform type, coefficient magnitude, prediction type, and boundary discontinuity, to improve efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single context is used for entropy coding of sign prediction errors across all transforms, then the device complexity is reduced, but the coding efficiency deteriorates
Solution Approach 1:
The patent applies local quality by selecting different entropy coding contexts based on the specific transform type (DST or DCT) and coefficient characteristics. Instead of using a uniform context for all transforms, the system adapts the context selection to local transform properties and coefficient magnitudes, thereby optimizing coding efficiency for each specific case while maintaining manageable complexity through rule-based selection.
2Productivity
If transform-specific contexts are used for entropy coding, then the coding efficiency is improved, but the device complexity increases
Solution Approach 1:
The patent implements parameter changes by dynamically selecting entropy coding contexts based on transform type parameters and coefficient magnitude parameters. The context selection adapts to changes in these parameters, allowing the system to optimize coding efficiency for different transform scenarios (DST vs DCT) and coefficient characteristics without requiring a complete redesign of the entropy coding framework.
3Productivity
If sign prediction is applied to all coefficients, then the coding efficiency is improved, but the processing complexity increases
Solution Approach 1:
The patent applies partial action by performing sign prediction selectively on significant coefficients rather than all coefficients. The system identifies coefficients that meet certain criteria (such as magnitude thresholds) and applies sign prediction only to those, thereby achieving coding efficiency improvements for the most impactful coefficients while avoiding the excessive processing complexity that would result from applying prediction to every coefficient in the block.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present disclosure provides embodiments for coding and decoding signs of transform coefficients which is applicable, for instance, in image and/or video coding and decoding. In particular, a plurality of the signs are predicted and only a prediction error signal is embedded in the bitstream. The prediction error signal may have a distribution which can be efficiently coded with CABAC or another variable length (entropy) coding. Moreover, if adaptive multi-core transform is used, the context for the entropy code to code a transform coefficient sign is selected according to the transformation which was used to obtain the transform coefficient.