Picture Prediction Method Using Neural Network Filtering for Video Coding

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing video coding technologies, such as H.266/VVC, do not fully consider the correlation between different video components, leading to reduced accuracy in inter-component prediction and decreased compression and coding efficiency due to discontinuity between prediction values and reference samples.

Innovation Solution

A picture prediction method that determines an initial prediction value for a colour component, uses side information from a reference component, and filters it through a preset network model to enhance the prediction value, improving accuracy and efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If existing prediction algorithms are used, then the prediction process is simple, but the prediction accuracy is reduced due to not fully considering correlation between different components

Engineering Contradiction:

Improveprediction accuracyVSAvoidprediction process complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

A neural network model is introduced as an intermediary component to process the correlation information between different colour components. The neural network receives input data about the correlation relationships and outputs enhanced prediction values, thereby improving prediction accuracy without requiring complex manual correlation processing in the traditional sense.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The prediction method changes the parameters used in the prediction process by incorporating correlation information between different colour components (e.g., luma and chroma components) into the prediction model. This allows the system to leverage statistical relationships between components to improve prediction accuracy while maintaining a relatively simple overall structure.

Inventive Principle:

Principle #35Parameter changes

2Measurement precision

If Cross-component Linear Model (CCLM) prediction mode is used, then the prediction process is standardized, but discontinuity occurs between prediction values and reference samples reducing prediction accuracy

Engineering Contradiction:

Improveprediction accuracyVSAvoidcontinuity between prediction values and reference samples

Core Design Contradiction:

Measurement precisionVSStability of the object's composition

Solution Approach 1:

The neural network model processes correlation information from reference samples and provides feedback in the form of enhanced prediction values. This feedback mechanism allows the system to adjust the prediction values based on the actual correlation patterns observed in the reference samples, thereby eliminating discontinuities and improving smoothness between prediction values and reference samples.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS12262041B2Picture prediction method, encoder, decoder and computer storage medium

Publication Date: 2025.03.25 GUANGDONG OPPO MOBILE TELECOMMUNICATIONS CORP LTD

AI summary

An image prediction method includes: an initial prediction value of a to-be-predicted colour component of a current block is determined; a sample associated with a reference colour component of the current block is determined; side information of the current block is determined according to the sample associated with the reference colour component; the initial prediction value is filtered by using a preset network model and the side information of the current block, to obtain a target prediction value of the to-be-predicted colour component of the current block; and the to-be-predicted colour component of the current block is encoded according to the target prediction value.