GPM Split Mode Reordering for Higher Video Coding Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding standards like VVC struggle to achieve superior coding efficiency, and there is a need for improved methods to enhance the coding efficiency of geometric partitioning (GPM) mode in video encoding and decoding processes.
Innovation Solution
The proposed method involves reordering GPM split modes based on distortion costs and using an order table to determine a reordered list of GPM split modes, allowing for more efficient video encoding and decoding by selecting the optimal GPM split mode index and predictor.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If GPM split modes are processed in conventional order, then decoding complexity is reduced, but coding efficiency is insufficient
Solution Approach 1:
The patent applies preliminary action by pre-defining and storing the optimal GPM split mode order in an order table during the encoding phase. The encoder calculates distortion costs for all GPM split modes, determines the optimal ordering that minimizes bit rate, and stores this ordering in advance. During decoding, the decoder simply retrieves the pre-defined order from the table, avoiding complex real-time calculations while achieving optimal coding efficiency.
Solution Approach 2:
The patent changes the parameter ordering of GPM split modes based on distortion cost analysis. Instead of using a fixed conventional order, the patent dynamically adjusts the sequence of GPM split modes according to their calculated distortion costs, transforming the parameter arrangement to optimize coding efficiency. This parameter reordering allows the most promising modes to be evaluated first, reducing the average bit rate required to signal the selected mode.
2Manufacturing precision
If more GPM split modes are evaluated to improve coding efficiency, then video quality improves, but computational cost increases
Solution Approach 1:
The patent performs the computationally intensive distortion cost calculations and optimal ordering determination in advance during encoding, storing the results in an order table. This preliminary action shifts the computational burden to a one-time setup phase rather than requiring high computational power during real-time decoding, enabling thorough evaluation of all GPM split modes without sustained high computational cost.
Solution Approach 2:
The patent uses the pre-calculated optimal order from the order table as a template for decoding multiple video frames. Instead of recalculating distortion costs for each frame, the decoder copies and applies the pre-determined optimal ordering, significantly reducing computational cost while maintaining the ability to achieve high video quality through comprehensive mode evaluation.
Data Source
AI summary
Methods for video decoding and encoding, apparatuses and non-transitory storage media are provided. In one decoding method, the decoder obtains a plurality of prediction samples of neighboring reconstructed samples of a coding unit (CU). Additionally, the decoder reorders a plurality of geometric partition mode (GPM) split modes based on distortion costs between the plurality of prediction samples of the neighboring reconstructed samples associated with each GPM split mode and neighboring reconstructed samples of the CU according to an order table to obtain a reordered list of the plurality of GPM split modes. Furthermore, the decoder obtains a GPM split mode index and then obtains a GPM split mode based on the GPM split mode index and the reordered list of the plurality of GPM split modes. Moreover, the decoder obtains a GPM predictor based on the GPM split mode.