CPB Removal Time Derivation for Video Decoding Error Resilience
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding methods are vulnerable to errors and inefficient in signaling and deriving coded picture buffer (CPB) removal times, as they rely on timing information from previous access units, leading to potential loss of data and unnecessary duplication of parameters.
Innovation Solution
The techniques provide a method to determine CPB removal times for a decoding unit independently of other access units, utilizing sub-picture timing SEI messages and a sequence level flag to control the presence of CPB parameters, expanding the definition of decoding units, and restricting buffering period SEI messages to improve error resilience and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If timing information from previous access units is used to derive CPB removal times, then signaling overhead is reduced, but error resilience deteriorates due to potential loss of timing information
Solution Approach 1:
The patent segments the timing information dependency by introducing independent timing references at different levels (picture-level and sub-picture-level). Instead of relying solely on previous access unit timing, each picture or sub-picture can have its own timing parameters, breaking the chain of dependency and improving error resilience while maintaining signaling efficiency through selective use of segmentation.
2Measurement precision
If CPB parameters are signaled for every access unit, then accuracy of timing information is improved, but bandwidth consumption increases
Solution Approach 1:
The patent applies local quality by allowing CPB parameters to be signaled at different granularities depending on the specific needs of each picture or sub-picture. Rather than uniformly signaling for all access units, the system selectively signals parameters where needed (at picture-level or sub-picture-level), achieving accurate timing information where necessary while minimizing unnecessary bandwidth consumption in other areas.
3Loss of information
If timing derivation relies on previous access units, then parameter duplication is reduced, but vulnerability to errors increases
Solution Approach 1:
The patent implements preliminary action by establishing independent timing references (picture-level and sub-picture-level CPB parameters) in advance, before potential errors can propagate. This allows the system to have pre-established timing information that does not depend on previous access units, reducing vulnerability to errors while avoiding unnecessary parameter duplication through the strategic placement of these preliminary timing references.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A video coding device, such as a video decoder, may be configured to derive at least one of a coded picture buffer (CPB) arrival time and a CPB nominal removal time for an access unit (AU) at both an access unit level and a sub-picture level regardless of a value of a syntax element that defines whether a decoding unit (DU) is the entire AU. The video coding device may further be configured to determine a removal time of the AU based at least in part on one of the CPB arrival time and a CPB nominal removal time and decode video data of the AU based at least in part on the removal time.