Entropy Coding Context Initialization With Reduced-Precision States
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current context-adaptive binary entropy coding methods for video data face challenges in achieving a balance between compression efficiency and implementation complexity, particularly in initializing probability estimations for contexts, which can lead to suboptimal coding efficiency due to high memory demands and complexity.
Innovation Solution
The proposed solution involves a decoder and encoder configuration that distinguishes between 126 probability states, initializing probability states using a linear equation of the quantization parameter, with the entropy decoder and encoder deriving slope and offset values from specific bits of an 8-bit initialization value, and updating probability states based on previously decoded or encoded portions of the data stream.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If high precision slope and offset values are used for initializing probability states, then coding efficiency is improved, but memory demands and implementation complexity increase
Solution Approach 1:
The patent applies this principle by using reduced precision (4-bit) slope and offset values instead of full precision (8-bit or higher) values. The lower precision values consume less memory and require simpler data structures, while the patent compensates for the reduced precision through context-adaptive selection and multiple candidate values, achieving acceptable coding efficiency with reduced complexity
Solution Approach 2:
The patent changes the precision parameter of the initialization values from high precision to reduced precision (4-bit values). It also introduces dynamic parameter selection by choosing different slope and offset values based on context conditions, allowing the system to adapt precision requirements to specific coding scenarios rather than uniformly using high precision throughout
2Measurement precision
If the number of contexts is increased, then probability estimation accuracy is improved, but adaptation fails due to low symbol frequency within individual contexts
Solution Approach 1:
The patent applies local quality by assigning different initialization characteristics (different slope and offset values) to different contexts based on their specific symbol statistics. Each context is customized with initialization parameters suited to its local characteristics rather than using a uniform initialization scheme, improving probability estimation accuracy for each specific context type
Solution Approach 2:
The patent performs preliminary initialization of probability states using context-dependent slope and offset values before actual coding begins. This preliminary action sets up appropriate starting conditions for each context based on expected symbol statistics, enabling faster and more reliable adaptation during the coding process
Data Source
AI summary
A decoder includes an entropy decoder configured to derive a number of bins of the binarizations from the data stream using binary entropy decoding by selecting a context among different contexts and updating probability states associated with the different contexts, dependent on previously decoded portions of the data stream; a desymbolizer configured to debinarize the binarizations of the syntax elements to obtain integer values of the syntax elements; a reconstructor configured to reconstruct the video based on the integer values of the syntax elements using a quantization parameter, wherein the entropy decoder is configured to distinguish between 126 probability states and to initialize the probability states associated with the different contexts according to a linear equation of the quantization parameter, wherein the entropy decoder is configured to, for each of the different contexts, derive a slope and an offset of the linear equation from first and second four bit parts of a respective 8 bit initialization value.