Adaptive Motion Compensated Filtering for Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding techniques, such as those based on the VVC standard, face limitations in achieving superior coding efficiency for inter coding blocks, particularly in reducing bit-rate while maintaining video quality.
Innovation Solution
The proposed method for video decoding and encoding of inter coding blocks involves obtaining prediction blocks and a current template, then using these to derive filters for improving prediction accuracy. This includes calculating target templates, minimizing differences to obtain filter coefficients, and applying these filters to prediction blocks to generate filtered blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional video coding techniques (VVC standard) are used for inter coding blocks, then coding efficiency is maintained at current levels, but bit-rate reduction and video quality improvement are limited
Solution Approach 1:
The patent applies preliminary action by performing adaptive filtering on prediction blocks before they are used for reconstruction. The filtering process is prepared in advance by deriving filter coefficients from template samples and applying them to motion-compensated prediction blocks, thereby improving prediction accuracy before the final reconstruction step occurs.
Solution Approach 2:
The patent implements parameter changes by dynamically adjusting filter coefficients based on local image characteristics. The filter coefficients are derived from template samples and adapted to match the specific content being coded, allowing the filtering parameters to change according to the local variance and characteristics of different video regions.
2Productivity
If adaptive filtering is applied to improve prediction accuracy, then coding efficiency increases, but computational complexity increases
Solution Approach 1:
The patent applies local quality by performing adaptive filtering selectively based on local image characteristics. The filtering is applied to regions where it provides benefit, with filter coefficients derived from local template samples. This allows coding efficiency to be improved in regions where adaptive filtering is most effective while avoiding unnecessary computation in regions where simple prediction suffices.
Solution Approach 2:
The patent implements partial action by applying adaptive filtering to specific prediction blocks where it provides the most benefit. Rather than uniformly applying filtering to all blocks, the method selectively applies filtering based on local characteristics, thereby improving coding efficiency for the most beneficial cases while limiting overall computational complexity.
3Measurement precision
If filter coefficients are derived from template samples, then prediction accuracy improves, but processing time increases
Solution Approach 1:
The patent applies preliminary action by deriving filter coefficients from template samples in advance, before the actual prediction and reconstruction steps. The template samples are collected and processed to determine filter coefficients that will be used for filtering prediction blocks, thereby preparing the filtering parameters beforehand to speed up the overall processing pipeline.
Solution Approach 2:
The patent implements parameter changes by dynamically adjusting filter coefficients based on the derived template samples. The coefficients are changed to match the local characteristics of each region, improving prediction precision while the changes are computed efficiently using the pre-derived template information.
Data Source
AI summary
Methods for video decoding and encoding, apparatuses and non-transitory computer-readable storage media thereof are provided. In one method for video decoding, a decoder may obtain a plurality of prediction blocks based on a current inter coding block; obtain a current template of the current inter coding block, wherein the current template comprises a plurality of reconstructed samples neighboring to the current inter coding block; obtain a plurality of template predictions of the current template respectively corresponding to the plurality of prediction blocks of the current inter coding block; obtain at least one filter based on the plurality of template predictions and the current template; and obtain a filtered block based on the at least one filter and the plurality of prediction blocks.