Template-Based Intra Prediction for Lower Video Signaling Overhead

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing video encoding and decoding technologies face challenges in efficiently reducing redundant information and signaling overhead, leading to increased data size and resource requirements for video storage and transmission.

Innovation Solution

The use of template-based determination of intra prediction modes, where weights for predicting block information are selected based on neighboring samples, allowing encoders to avoid signaling these weights, thereby improving coding efficiency and reducing signaling overhead.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If weights are separately signaled for decoding blocks, then decoding accuracy is improved, but signaling overhead is increased

Engineering Contradiction:

Improvedecoding accuracyVSAvoidsignaling overhead

Core Design Contradiction:

Measurement precisionVSQuantity of substance

Solution Approach 1:

The encoder determines weights based on neighboring samples without requiring separate signaling of these weights. The decoder independently calculates the same weights using the same neighboring samples, allowing the system to serve itself without additional signaling overhead for weight transmission.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

Neighboring samples serve as an intermediary that both encoder and decoder use to derive weights. Instead of directly transmitting weights, the system uses neighboring samples as a common reference point that mediates the weight determination process for both encoding and decoding operations.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Productivity

If more information is encoded for video compression, then coding efficiency is improved, but data size increases

Engineering Contradiction:

Improvecoding efficiencyVSAvoiddata size

Core Design Contradiction:

ProductivityVSQuantity of substance

Solution Approach 1:

The patent extracts and eliminates the need to signal weights separately from the bitstream. By deriving weights from neighboring samples that are already available in the decoded data, the system removes unnecessary signaling elements while maintaining coding efficiency through template-based weight determination.

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS12457328B2Video compression using template-based determination of intra prediction mode

Publication Date: 2025.10.28 COMCAST CABLE COMM LLC

AI summary

Encoding and/or decoding a block of a video frame may be based on previously decoded reference information in the frame. A weighted sum of the reference information may be used as a prediction of the block for encoding and/or decoding the block. The weights to be applied for determining the weighted sum may be based on weights that most accurately predict neighboring, template samples of the block.