Significance Map Context Sharing Across Multiple Transform Sizes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding standards, such as MPEG-4 AVC and KTA, face challenges in efficiently coding significance maps for transform coefficients across different transform sizes, leading to inadequate capture of coefficient distributions and complexity in extending designs to future standards with more transforms.
Innovation Solution
A unified rule-based approach is proposed for generating context sharing maps that adapt to transform sizes, allowing multiple transform coefficient positions to share contexts, simplifying encoder and decoder design and enabling easy extension to future standards.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If separate context models are used for different transform sizes, then the coding accuracy for each transform is improved, but the device complexity and number of context models increases
Solution Approach 1:
The patent creates a unified context model that serves multiple transform sizes (4x4, 8x8, 16x16, and future transform sizes) simultaneously. Instead of having separate context models for each transform size, a single context model is designed to adaptively handle coefficients from different transform sizes through position-based context selection, reducing the total number of context models while maintaining coding accuracy
Solution Approach 2:
The patent changes the parameter of context model selection from transform-size-specific to position-based. By using a unified rule that maps coefficient positions to context models regardless of transform size, the system maintains adaptability to different transform sizes while using a fixed, manageable number of context models
2Productivity
If more context models are created to capture coefficient distributions for different transforms, then the coding efficiency is improved, but the ease of manufacture and extension to future standards deteriorates
Solution Approach 1:
The unified context model design allows the same context model to be used across current and future transform sizes. The model structure is designed to be transform-size-agnostic, depending only on coefficient position, which simplifies the extension process to future standards while maintaining high coding efficiency through adaptive context selection
Solution Approach 2:
The patent segments the context modeling approach by position rather than by transform size. This segmentation strategy allows the same context model to be divided into position-specific contexts that work universally across different transform sizes, maintaining coding efficiency without increasing design complexity
3Measurement precision
If transform size-specific context sharing maps are used, then the capture of coefficient distributions is improved, but the adaptability to future transforms with different sizes deteriorates
Solution Approach 1:
The patent changes the fundamental parameter for context sharing from transform size to coefficient position within the transform. This parameter change makes the context sharing rule transform-size-independent, allowing the same unified rule to work for current and future transform sizes while accurately capturing coefficient distributions through position-based context assignment
Data Source
AI summary
Methods and apparatus are provided for unified significance map coding. An apparatus includes a video encoder for encoding transform coefficients for at least a portion of a picture. The transform coefficients are obtained using a plurality of transforms. One or more context sharing maps are generated for the transform coefficients based on a unified rule. The one or more context sharing maps are for providing at least one context that is shared among at least some of the transform coefficients obtained from at least two different ones of the plurality of transforms.