Reference Picture List Derivation for Low-Overhead Video Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in efficiently signaling reference picture lists, leading to inefficiencies in the coding process, particularly in scenarios with limited network resources and high video quality demands.
Innovation Solution
The proposed method involves parsing parameter sets in a coded video bitstream to derive reference picture lists, ensuring the order of entries in the list matches the order of corresponding reference pictures, and utilizing structures like RefPictList[0] or RefPictList[1], which simplifies and enhances the signaling of reference picture lists.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If complex reference picture list signaling methods are used, then reference picture management accuracy is improved, but coding complexity increases
Solution Approach 1:
The reference picture list is divided into multiple sub-lists (List 0 and List 1) with specific construction rules for each. This segmentation allows precise control over which reference pictures are used for different prediction modes while simplifying the overall signaling structure by assigning dedicated construction methods to each sub-list.
Solution Approach 2:
Reference pictures are pre-ordered and pre-organized into candidate lists before actual prediction coding. The lists are constructed in advance with specific ordering rules (e.g., temporal distance, picture order count) so that during decoding, the reference picture selection process is simplified and requires minimal additional signaling.
2Manufacturing precision
If more reference pictures are included in reference picture lists, then prediction accuracy is improved, but memory resources increase
Solution Approach 1:
The patent introduces configurable parameters such as max_num_refpictures and max_num_refpictures_list1 to control the maximum number of reference pictures in each list. These parameters allow dynamic adjustment of memory usage based on available resources while maintaining sufficient reference pictures for accurate prediction, enabling adaptation to different deployment scenarios.
Solution Approach 2:
The system constructs reference picture lists with a predetermined maximum number of entries even when fewer reference pictures are available. This partial filling approach ensures consistent list structures for efficient processing while avoiding the need to allocate memory for potentially unused reference picture slots, optimizing memory utilization.
Data Source
AI summary
A method of decoding a coded video bitstream is provided. The method includes parsing a parameter set represented in the coded video bitstream. The parameter set contains a set of syntax elements that include a set of reference picture list structures. The method also includes parsing a slice header of a current slice represented in the coded video bitstream. The slice header contains an index of a reference picture list structure among the set of reference picture list structures in the parameter set. The method further includes deriving, based on the set of syntax elements in the parameter set and the index of the reference picture list structure, a reference picture list of the current slice. The method also includes obtaining, based on the reference picture list, at least one reconstructed block of the current slice.