Parsing-Independent Transform Prediction Through Local Identifier Derivation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing image and video compression technologies face limitations in coding efficiency, particularly in signaling transform identifiers, which hinders accurate decoding and encoding processes.

Innovation Solution

The proposed method omits signaling transform identifiers by allowing decoders to derive them based on transform-specific probability distributions and quantized coefficients, enabling block-parsing-independent decoding and encoding.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If transform identifiers are signaled in the bitstream, then decoding accuracy is improved, but bitstream complexity and transmission bandwidth increase

Engineering Contradiction:

Improvedecoding accuracyVSAvoidbitstream complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts and removes the transform identifier signaling from the bitstream. Instead of transmitting transform identifiers explicitly, the decoder derives them locally using probability distributions and quantized coefficient patterns. This extraction eliminates redundant information transmission while maintaining decoding accuracy through local derivation mechanisms.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The decoder performs self-service by locally deriving transform identifiers using transform-specific probability distributions and quantized coefficient patterns available in the bitstream. This self-derivation mechanism eliminates the need for external signaling of transform identifiers, reducing bitstream complexity while maintaining decoding accuracy.

Inventive Principle:

Principle #25Self-service

2Measurement precision

If transform identifiers are signaled in the bitstream, then transform identification accuracy is improved, but transmission bandwidth is consumed

Engineering Contradiction:

Improvetransform identification accuracyVSAvoidtransmission bandwidth

Core Design Contradiction:

Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent extracts transform identifier information from the quantized coefficient patterns themselves rather than transmitting it separately. The transform identifiers are embedded in the coefficient data structure, allowing the decoder to extract them through local analysis without consuming additional transmission bandwidth.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent merges transform identifier information with the quantized coefficient data. Instead of separate signaling, the transform identification information is combined with the coefficient representation, allowing both transformation and identification functions to be achieved through a unified data structure that consumes less bandwidth.

Inventive Principle:

Principle #5Merging (Combining)

3Ease of operation

If transform identifiers are decoded before transform coefficients, then decoding sequence is simplified, but parsing dependency increases processing complexity

Engineering Contradiction:

Improvedecoding sequence simplicityVSAvoidparsing dependency

Core Design Contradiction:

Ease of operationVSDevice complexity

Solution Approach 1:

The patent inverts the traditional decoding sequence by deriving transform identifiers concurrently with or after coefficient decoding rather than before. This inversion eliminates parsing dependency because the identifier derivation uses information already available during coefficient decoding, allowing asynchronous and independent processing of transform data.

Inventive Principle:

Principle #13The other way round (Inversion)

Solution Approach 2:

The patent performs preliminary preparation of transform-specific probability distributions and coefficient patterns during encoding, so that the decoder can directly derive transform identifiers without sequential parsing dependencies. This preliminary action enables parallel and asynchronous decoding operations.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS12432383B2Transform prediction with parsing independent coding

Publication Date: 2025.09.30 GOOGLE LLC

AI summary

Transform prediction with parsing independent coding includes generating a reconstructed frame and outputting the reconstructed frame. Generating the reconstructed frame includes entropy decoding transform blocks for the reconstructed frame, entropy decoding decoded transform identifiers for the transform blocks, obtaining transform-specific probability distributions for available transforms, and, for a current transform block from the transform blocks, identifying a current remapped transform identifier from the decoded transform identifiers, identifying a current transform identifier in accordance with the current remapped transform identifier, the transform coefficients from the current transform block, and the transform-specific probability distributions, identifying a current transform in accordance with the current transform identifier; inverse transforming, in accordance with the current transform, the current transform block to obtain a current residual block and obtaining a current reconstructed block using the current residual block. Generating the reconstructed frame includes including the current reconstructed block in the reconstructed frame.