Decoded Picture Buffer Signaling for Multi-Layer Video Bitstreams
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding standards face challenges in efficiently managing and signaling decoded picture buffer parameters across multiple layers in scalable video coding, leading to inefficiencies in bandwidth usage and decoding processes.
Innovation Solution
Implementing a format rule for video processing that specifies how decoded picture buffer parameters are signaled and structured within the bitstream, including rules for profile-tier-level syntax structures, output layer sets, and reference decoder parameters, to optimize the encoding and decoding of videos with multiple layers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If the VVC scalability design uses existing constraints on slice types, then the bitstream format is simplified, but unnecessary constraints are imposed that do not align with the scalability structure
Solution Approach 1:
The patent modifies the slice type constraints by introducing a new parameter `slice_intra_only_flag` that can take values 0 or 1, allowing flexible configuration of intra-only slices within scalable video coding. This parameter change enables the bitstream format to accommodate scalability requirements without imposing unnecessary constraints, resolving the contradiction between format simplicity and constraint accuracy.
2Measurement precision
If all syntax elements are explicitly signaled, then parameter accuracy is maintained, but redundancy and unnecessary signaling increase
Solution Approach 1:
The patent extracts and removes unnecessary syntax elements from the bitstream by introducing a mechanism to infer certain parameters based on context. Specifically, when `slice_intra只有_flag` is not explicitly signaled, its value can be inferred from the parent slice header, eliminating redundant signaling while maintaining parameter accuracy. This extraction of unnecessary elements reduces bitstream redundancy without sacrificing measurement precision.
3Quantity of substance
If HRD parameters are repeated for each output layer set, then parameter completeness is ensured, but bitstream length and complexity increase
Solution Approach 1:
The patent merges HRD parameters across multiple output layer sets by introducing a flag `hrd_params_inferred` that allows certain HRD parameters to be inferred rather than explicitly repeated. When this flag is set, the decoder can derive HRD parameters for one layer set from another layer set's parameters, combining information across layers. This merging approach ensures parameter completeness while reducing bitstream length by eliminating redundant repetitions.
4Quantity of substance
If PTL syntax structures are included for all output layer sets, then profile information completeness is improved, but unnecessary repetition occurs for single-layer sets
Solution Approach 1:
The patent applies partial action by including PTL syntax structures only when necessary. The patent introduces a condition that allows the PTL syntax structure to be omitted for output layer sets that have only a single layer, as the profile information can be inferred from the base layer. This partial inclusion approach maintains profile information completeness for multi-layer sets while avoiding unnecessary repetition for single-layer sets, reducing syntax structure complexity.
Data Source
AI summary
A method of video processing includes performing a conversion between a video and a bitstream of the video. The bitstream includes one or more output layer sets each comprising one or more video layers. The bitstream conforms to a format rule, wherein the format rule specifies that a number of decoded picture buffer parameter syntax structures included in a video parameter set for the bitstream is equal to: zero, in a case that each output layer set includes a single video layer; or one plus a value of a syntax element, in a case that each output layer set includes a single layer is not true.