TCQ State Transition Tables for Higher-Precision Video Quantization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing video coding systems face challenges in efficiently compressing digital video signals, particularly in managing quantization states and dequantization indices, which affect compression efficiency and quality.

Innovation Solution

Implementing a trellis coded quantization (TCQ) scheme with a quantization state transition table that includes nine states, including eight quantization states and a dummy state, to determine dequantization indices based on parity and associate odd and even indices with different scalar quantizers, allowing for enhanced quantization and dequantization processes at block, slice, or picture levels.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a quantization state transition table with more than eight states is used for trellis coded quantization, then compression efficiency and quality are improved, but device complexity and processing overhead increase

Engineering Contradiction:
Improvequantization precisionVSAvoidquantization state management complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The quantization state transition table is segmented into multiple states (more than eight states) to provide finer quantization granularity. This segmentation allows the system to represent quantization levels with higher precision while managing complexity through structured state transitions defined in the table.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The invention changes the parameter of quantization state count from the conventional eight states to more than eight states. This parameter change enables enhanced rate-distortion performance and compression efficiency by providing more quantization levels, while the dummy state mechanism helps manage the increased complexity.

Inventive Principle:
Principle #35Parameter changes

2Adaptability or versatility

If multiple dequantization indices are associated with multiple quantization states, then quantization flexibility and rate-distortion performance are improved, but processing complexity and computational overhead increase

Engineering Contradiction:
Improvequantization adaptabilityVSAvoiddequantization index management complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The quantization state transition table serves multiple functions: it defines state transitions, associates dequantization indices with quantization states, and provides a framework for selecting appropriate quantization states. This multi-functionality improves quantization adaptability while avoiding the need for separate management mechanisms for each function.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The dummy state acts as an intermediary element in the quantization state transition table. It facilitates the association between quantization states and dequantization indices without requiring direct complex mapping, thereby simplifying the management of multiple dequantization indices while maintaining quantization flexibility.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Ease of operation

If a dummy state is added to the quantization state transition table, then state transition management and dequantization index association are simplified, but the table size and memory requirements increase

Engineering Contradiction:
Improvestate transition management easeVSAvoidquantization state transition table size
Core Design Contradiction:
Ease of operationVSQuantity of substance

Solution Approach 1:

The dummy state is extracted as a separate, dedicated element within the quantization state transition table. This extraction allows the dummy state to serve its special function of simplifying state transition management and dequantization index association without interfering with the normal quantization states, while its impact on table size is minimized through efficient encoding.

Inventive Principle:
Principle #2Taking out (Extraction)

4Productivity

If quantization is performed at block, slice, or picture level with enhanced state transition tables, then compression efficiency is improved, but computational complexity and processing time increase

Engineering Contradiction:
Improvecompression efficiencyVSAvoidquantization processing time
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The quantization state transition table is prepared and configured in advance with all necessary state transitions and dequantization index associations. This preliminary action allows the encoding process to simply look up pre-computed values during quantization, improving compression efficiency while minimizing the computational time required during actual video encoding.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4668741A1Enhanced states dependent quantization
Publication Date: 2025.12.24 INTERDIGITAL CE PATENT HOLDINGS SAS
  • EP4668741A1 patent drawingFigure 1A
  • EP4668741A1 patent drawingFigure 1B
  • EP4668741A1 patent drawingFigure 1C

AI summary

Systems, methods, and instrumentalities may be provided for enhanced states dependent quantization. In examples, a device (e.g., a video encoding device) may determine to use a quantization state transition table for trellis coded quantization (TCQ) comprising a plurality of quantization states and a dummy state. The device may determine that the quantization state transition table comprises more than eight states. The device may select a quantization state from the plurality of quantization states. The device may determine, based on the selected quantization state, a dequantization index based on a parity associated with the selected quantization state. The dummy state may be associated with the selected quantization state. The device may perform quantization based on the selected quantization state.