Video Coding OLS PTL Inference and Conformance Window

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current Versatile Video Coding (VVC) high-level syntax (HLS) design has limitations that can be improved to reduce bitrate without compromising video fidelity, particularly in signaling correspondences between output layer sets (OLSs) and profile-tier-level (PTL) data structures, and in inferring conformance cropping window data from sequence parameter sets (SPS) when picture sizes are maximum.

Innovation Solution

The techniques involve inferring correspondences between OLSs and PTL data structures when their numbers are equal, avoiding explicit signaling of dependency information for each layer, and simplifying the signaling of conformance cropping window data by inferring it from the SPS when picture sizes are maximum, thereby reducing processing cycles and bitrate.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If explicit signaling of correspondence between OLSs and PTL data structures is performed, then reliability of parameter mapping is improved, but signaling overhead increases

Engineering Contradiction:
Improveparameter mapping reliabilityVSAvoidsignaling overhead
Core Design Contradiction:
ReliabilityVSLoss of information

Solution Approach 1:

The system uses the equality between the number of OLSs and PTL data structures as a self-indicating mechanism. When these numbers are equal, the correspondence is automatically established without needing additional signaling, allowing the data structure itself to indicate its own mapping relationship

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent extracts the correspondence indication function from separate signaling elements and integrates it into the existing number field. The same field that indicates the number of OLSs/PTLs also serves to indicate their correspondence when the values are equal, eliminating redundant information transmission

Inventive Principle:
Principle #2Taking out (Extraction)

2Manufacturing precision

If conformance cropping window data is explicitly signaled in PPS, then precision of conformance control is improved, but processing overhead increases

Engineering Contradiction:
Improveconformance control precisionVSAvoidprocessing overhead
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The SPS conformance cropping window data structure is designed to serve multiple purposes: it provides conformance control for both SPS-level and PPS-level operations. By making the SPS data universal, the patent eliminates the need for separate PPS signaling while maintaining precise conformance control across different parameter set levels

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The conformance cropping window data is established at the SPS level in advance, before PPS-specific processing occurs. This preliminary establishment of conformance parameters allows PPS to inherit and use these pre-defined values without needing to re-signal them, reducing processing overhead

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS11778215B2Coding output layer set data and conformance window data of high level syntax for video coding
Publication Date: 2023.10.03 QUALCOMM INC
  • US11778215B2 patent drawing
  • US11778215B2 patent drawing
  • US11778215B2 patent drawing

AI summary

In one example, a device for decoding video data includes one or more processors implemented in circuitry and configured to: determine that a value for a syntax element representing a number of profile-tier-level (PTL) data structures in a video parameter set (VPS) of a bitstream is equal to a total number of output layer sets (OLSs) specified for the VPS; in response to determining that the value for the syntax element representing the number of profile-tier-level data structures in the VPS is equal to the total number of OLSs specified for the VPS, infer values for OLS PTL index values, without explicitly decoding values for the OLS PTL index values; and decode video data of one or more OLSs using corresponding PTL data structures of the PTL data structures in the VPS according to the inferred values for the OLS PTL index values.