Transform Coefficient Coding Using Shared Context and Symbol Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing transform coefficient coding schemes face challenges in maintaining low complexity while achieving high coding efficiency, especially with variable block sizes and the need to convey additional data such as depth maps and transparity values, which increases the number of contexts and requires precise adaptation of symbolization schemes to coefficient statistics.
Innovation Solution
A coding apparatus that maps transform coefficients to symbols using parameterizable symbolization schemes and context adaptive entropy encoding, where the same function is used for context dependency and symbolization parameter determination, allowing for efficient processing across different block sizes and frequency portions, and inserting symbols into a data stream based on previously coded coefficients.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If multiple different contexts are used to precisely estimate the probability of transform coefficient levels, then coding efficiency is improved, but device complexity increases due to the increasing number of different contexts with different functions
Solution Approach 1:
The patent changes the parameter of context identification from using multiple different context functions to using a single context function with adjustable parameters. The context is determined by a function of previously coded transform coefficients, where the function parameters can be adapted to different block sizes and coefficient types, replacing the need for multiple fixed context functions.
Solution Approach 2:
The patent creates a universal context determination function that can handle multiple different coding scenarios (different block sizes, different coefficient types) through parameter adjustment rather than requiring separate context functions for each case. This single function serves multiple purposes that previously required multiple specialized functions.
2Loss of information
If various factors are taken into account to adapt the symbolization scheme closely to the actual statistics, then coding efficiency is improved, but device complexity increases due to the huge amount of differing symbolization schemes required
Solution Approach 1:
The patent uses a parameterizable symbolization scheme where the binarization method (truncated unary, truncated binary, or Rice binarization) and its parameters can be selected and adjusted based on the transform coefficient level and position. This allows close adaptation to actual statistics through parameter selection rather than requiring separate symbolization schemes for each case.
Solution Approach 2:
The patent makes the symbolization scheme dynamic by allowing the binarization method and parameters to be selected adaptively based on the current transform coefficient's level and position, rather than using fixed symbolization schemes. This dynamic adaptation achieves close matching to actual statistics without requiring a huge number of predefined schemes.
Data Source
Figure 1
Figure 2~3
Figure 4
AI summary
An idea used herein is to use the same function for the dependency of the context and the dependency of the symbolization parameter on previously coded/decoded transform coefficients. Using the same function - with varying function parameter - may even be used with respect to different transform block sizes and/or frequency portions of the transform blocks in case of the transform coefficients being spatially arranged in transform blocks. A further variant of this idea is to use the same function for the dependency of a symbolization parameter on previously coded/decoded transform coefficients for different sizes of the current transform coefficient's transform block, different information component types of the current transform coefficient's transform block and/or different frequency portions the current transform coefficient is located within the transform block.