Intra Template Prediction with Candidate BVs for Video Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing digital video compression standards face challenges in improving compression efficiency, particularly in handling high-definition video content, where existing methods struggle to effectively utilize spatial and temporal correlations for accurate prediction and filtering.
Innovation Solution
The introduction of a decoding method that utilizes intra template prediction with candidate block vectors (BVs) and a first identifier and index to determine a predicted block, along with an encoding method that encodes these parameters, enhances the accuracy of prediction and improves decoding performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If existing digital video compression standards are used, then video transmission and storage are facilitated, but compression efficiency is insufficient for high-definition video content
Solution Approach 1:
The patent changes the parameters of template matching by introducing a filtering mechanism that processes candidate block vectors through a learning model. This filtering operation modifies how candidate blocks are selected and weighted, improving prediction accuracy while maintaining compression efficiency through optimized parameter usage in the prediction process
2Measurement precision
If traditional intra prediction methods are used, then decoding is simpler, but prediction accuracy for high-definition video is insufficient
Solution Approach 1:
The patent applies preliminary action by pre-training a learning model offline to generate filtering coefficients. These pre-computed coefficients are then used during decoding without requiring complex real-time learning operations, thus improving prediction accuracy while keeping decoding complexity manageable through the use of pre-processed data
Solution Approach 2:
The patent introduces an intermediary filtering mechanism that processes candidate block vectors before final prediction. This intermediary step uses pre-trained model coefficients to weight and select candidates, bridging the gap between simple traditional methods and complex deep learning approaches, thereby improving accuracy without proportionally increasing decoding complexity
Data Source
AI summary
A decoding method, an encoding method, a decoder, and an encoder are provided. The decoding method includes: obtaining a first identifier for indicating whether filtering is performed and a first index; determining, based on a first prediction mode using intra template prediction, a first candidate list obtained based on candidate block vectors (BVs) of a current block; and determining a predicted block of the current block based on a candidate BV indicated by the first index in the first candidate list and the first identifier.