NN-Based Intra Prediction Signaling for Lower Video Decoding Complexity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing NN-based intra-prediction in video coding faces inefficiencies in signaling the usage of the mode within the video coding framework, leading to disruptions in the coding pipeline and increased decoding complexity.
Innovation Solution
The signaling of the NN-intra mode flag is moved to the coding unit (CU) prediction data, following decoder-side intra mode derivation (DIMD), ensuring better integration with the coding pipeline and improving compression performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If the NN-intra mode flag is signaled before the DIMD flag, then the signaling is completed early in the bitstream, but the coding pipeline is disrupted and decoding complexity increases
Solution Approach 1:
The patent applies preliminary action by moving the NN-intra mode flag signaling to occur after the DIMD flag, allowing the decoder to first establish the decoding context and mode derivation framework before encountering the neural network mode indicator. This sequencing prevents disruption to the coding pipeline while maintaining efficient bitstream processing.
2Productivity
If the NN-intra mode flag is signaled before the DIMD flag, then compression efficiency may be improved, but the coding pipeline integration is suboptimal
Solution Approach 1:
The patent applies dynamics by adapting the signaling position of the NN-intra mode flag to match the dynamic requirements of the coding pipeline. By placing the flag after the DIMD flag, the signaling mechanism dynamically adjusts to the decoding context establishment phase, ensuring optimal integration with the coding pipeline while maintaining compression efficiency.
Data Source
AI summary
A video coder may code a flag indicating whether or not a neural network (NN)-based intra prediction mode is enabled before coding a decoder side intra mode derivation (DIMD) flag, and code a block of video data based on whether or not the NN-based intra prediction mode is enabled.