Decoder-Side Intra Mode Derivation Using Template Signaling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in efficiently deriving intra prediction modes, leading to suboptimal compression and quality degradation in video data transmission and storage.
Innovation Solution
Implementing template-based intra mode derivation (TIMD) methods at the decoder side, utilizing signaled flags to determine the application of TIMD mode and derive intra prediction modes based on templates or reference lines for current video blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If traditional intra prediction mode derivation is used, then device complexity is reduced, but video quality and compression performance deteriorate
Solution Approach 1:
The decoder autonomously derives the intra prediction mode by selecting a template from multiple candidate templates and determining the prediction mode based on template matching, without requiring explicit mode signaling from the encoder. This self-service mechanism improves video quality while minimizing additional complexity by reusing existing decoded pixel data.
Solution Approach 2:
The method performs template selection and mode determination in advance during the decoding process before final prediction, using previously decoded neighboring block data. This preliminary action allows the system to prepare optimal prediction parameters without adding significant computational overhead during critical path operations.
2Measurement precision
If more template candidates are considered for TIMD, then prediction accuracy improves, but processing time and complexity increase
Solution Approach 1:
The method considers a limited set of candidate templates (e.g., left, top, top-left, top-right neighbors) rather than exhaustively searching all possible templates. This partial action approach achieves sufficient prediction accuracy by focusing on the most relevant neighboring regions, thereby avoiding excessive processing time while maintaining adequate prediction quality.
3Productivity
If TIMD mode is applied to all video blocks, then compression performance improves, but bit rate increases due to additional signaling
Solution Approach 1:
The method applies TIMD selectively to specific video blocks where template-based prediction is beneficial, rather than uniformly to all blocks. By evaluating local characteristics and applying TIMD only where appropriate, the system improves compression performance for suitable blocks while avoiding unnecessary bit rate increases for blocks where traditional prediction suffices.
Data Source
AI summary
Methods, apparatuses, and non-transitory computer-readable storage mediums are provided for video coding. In one method, a decoder obtains a first signaled flag that indicates whether a template-based intra mode derivation (TIMD) mode is applied on a current video block; and in response to determining that the TIMD mode is applied on the current video block based on the first signaled flag, the decoder obtains a second signaled flag in transform block (TB), coding block (CB), slice, picture, or sequence level that indicates whether signaling of an index of a template for TIMD is enabled, where the index of the template is signaled to derive the intra prediction mode for the current video block.