Context-Adaptive Binary Arithmetic Coding Probability Update

Solution Overview

Problem

Context-adaptive binary arithmetic coding (CABAC) in video data encoding faces challenges due to high computational complexity and memory requirements from numerous contexts needed to capture statistical variations, limiting hardware implementability and throughput.

Innovation Solution

A method that reduces the number of contexts by using a single context for each binary symbol, modifying its probability based on previously coded symbols, and updating the context value after encoding, thereby simplifying the arithmetic coding process without significant loss of compression performance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of information

If numerous contexts are used to capture statistical variations in CABAC, then compression performance is improved, but device complexity and memory requirements increase

Engineering Contradiction:

Improvecompression performanceVSAvoidnumber of contexts

Core Design Contradiction:

Loss of informationVSDevice complexity

Solution Approach 1:

The patent merges multiple context models into a single unified context model. Instead of maintaining separate context tables for different syntax elements and positions, the invention uses one context model that adapts dynamically to different coding situations through probability modification based on previously coded symbols, thereby reducing memory requirements while preserving compression performance.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The patent introduces dynamic probability modification to the context model. The probability of the current binary symbol is modified based on the values of previously coded binary symbols, allowing the single context model to adapt to different statistical variations dynamically rather than requiring multiple static context tables.

Inventive Principle:

Principle #15Dynamics

2Loss of information

If numerous contexts are used to capture statistical variations in CABAC, then compression performance is improved, but hardware implementability decreases

Engineering Contradiction:

Improvecompression performanceVSAvoidhardware implementability

Core Design Contradiction:

Loss of informationVSEase of manufacture

Solution Approach 1:

The patent merges multiple context models into a single unified context model. Instead of maintaining separate context tables for different syntax elements and positions, the invention uses one context model that adapts dynamically to different coding situations through probability modification based on previously coded symbols, thereby reducing memory requirements while preserving compression performance.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The patent introduces dynamic probability modification to the context model. The probability of the current binary symbol is modified based on the values of previously coded binary symbols, allowing the single context model to adapt to different statistical variations dynamically rather than requiring multiple static context tables.

Inventive Principle:

Principle #15Dynamics

3Loss of information

If numerous contexts are used to capture statistical variations in CABAC, then compression performance is improved, but throughput decreases

Engineering Contradiction:

Improvecompression performanceVSAvoidencoding throughput

Core Design Contradiction:

Loss of informationVSProductivity

Solution Approach 1:

The patent merges multiple context models into a single unified context model. Instead of maintaining separate context tables for different syntax elements and positions, the invention uses one context model that adapts dynamically to different coding situations through probability modification based on previously coded symbols, thereby reducing memory requirements while preserving compression performance.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The patent introduces dynamic probability modification to the context model. The probability of the current binary symbol is modified based on the values of previously coded binary symbols, allowing the single context model to adapt to different statistical variations dynamically rather than requiring multiple static context tables.

Inventive Principle:

Principle #15Dynamics

Data Source

PatentUS11323714B2Method and device for context-adaptive binary arithmetic coding a sequence of binary symbols representing a syntax element related to video data

Publication Date: 2022.05.03 INTERDIGITAL VC HOLDINGS INC

AI summary

The present principles relate to a method and device for context-adaptive binary arithmetic coding a sequence of binary symbols representing a syntax element related to video data or a syntax element relative to a video data. The method comprises, for each binary symbol of the sequence of binary symbols:—obtaining (100) a context value from a context model defined for the binary symbol, said context value comprising bits representing the probability, called a first probability p, for the binary symbol to be equal to a binary value;—determining (110) a second probability p′ by modifying said first probability p according to at least one previously coded binary symbol of the sequence of binary symbols;—arithmetic coding (120) the binary symbol based on said second probability p′; and—updating and storing (130) the first probability p of said context value