Gradual Decoding Refresh for Low-Latency Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in achieving efficient compression and reduced latency for video streaming and storage, particularly in low-delay applications, due to the use of intra random access point (IRAP) pictures, which incur higher bitrates and increased end-to-end delay.
Innovation Solution
The implementation of gradual decoding refresh (GDR) techniques, using a GDR flag to initiate decoding at a GDR picture instead of an IRAP picture, allowing progressive intra refresh without the need for IRAP, thereby reducing latency and achieving a smoother bitrate.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If intra random access point (IRAP) pictures are used to enable random access in video coding, then random access capability is improved, but bitrate and end-to-end delay increase
Solution Approach 1:
The patent segments the picture structure into refresh regions and non-refresh regions, allowing incremental updates rather than requiring complete IRAP pictures. This segmentation enables random access points to be established more frequently with smaller picture sizes, reducing delay while maintaining adaptability.
Solution Approach 2:
The patent introduces dynamic control mechanisms including a GDR flag and recovery point counters that allow the video coding system to adaptively switch between different access modes. This dynamic approach enables the system to provide random access capability on demand without permanently increasing bitrate or delay for all pictures.
2Reliability
If intra random access point (IRAP) pictures are used to ensure decoding reliability, then decoding reliability is improved, but bitrate increases
Solution Approach 1:
The patent applies partial action by implementing gradual decoding refresh that updates only specific regions of pictures rather than requiring complete IRAP pictures. This allows decoding reliability to be maintained for critical regions while using fewer bits overall, as non-critical regions can be reconstructed from previous frames.
Solution Approach 2:
The patent changes the parameter of picture completeness by introducing recovery point pictures that are incomplete but sufficient for random access. By changing from complete IRAP pictures to partial recovery point pictures with GDR flags, the system maintains decoding reliability while reducing bitrate requirements.
3Loss of time
If progressive intra refresh is implemented to reduce latency, then end-to-end delay is reduced, but bitrate consistency becomes challenging
Solution Approach 1:
The patent implements feedback mechanisms through recovery point counters and GDR flags that monitor the decoding process and provide control information back to the encoder. This feedback allows the system to adjust the refresh rate and picture structure to maintain bitrate consistency while achieving reduced latency through progressive intra refresh.
Solution Approach 2:
The patent ensures continuity of useful action by implementing gradual decoding refresh that continuously updates picture regions without interrupting the video stream. This continuous refresh process maintains bitrate consistency by distributing the refresh overhead across multiple pictures rather than concentrating it in single IRAP pictures, while simultaneously reducing end-to-end delay.
Data Source
AI summary
A method of decoding a coded video bitstream implemented by a video decoder is provided. The method includes the video decoder receiving the coded video bitstream, the coded video bitstream containing a gradual decoding refresh (GDR) flag corresponding to a coded video sequence (CVS); determining, by the video decoder, whether a GDR picture is present in the CVS based on a value of the GDR flag; initiating, by the video decoder, decoding of the CVS at the GDR picture when the value of the GDR flag indicates that the GDR picture is present; and generating, by the video decoder, an image according to the CVS as decoded. A corresponding method of encoding implemented by a video encoder is also disclosed.