SEI Message for Decoder Picture-Based Parallelization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current video decoding technologies require deep analysis of bitstreams to identify independently decodable pictures, which is inefficient and hinders parallelization of decoding processes.

Innovation Solution

Incorporating a Supplementary Enhancement Information (SEI) message that indicates picture independence, allowing decoders to identify and decode pictures independently without requiring deep bitstream analysis.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If deep bitstream analysis is performed to identify independently decodable pictures, then decoding accuracy is improved, but decoding speed and parallelization capability deteriorate

Engineering Contradiction:
Improvedecoding accuracyVSAvoiddecoding speed
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The SEI message is inserted into the bitstream in advance to declare picture independence information. This allows the decoder to know beforehand which pictures can be decoded independently, eliminating the need for deep analysis during decoding and enabling immediate parallelization decisions.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The SEI message acts as an intermediary that carries picture independence information from the encoder to the decoder. This intermediary structure provides a standardized, easily parseable mechanism for conveying decoding dependency information without requiring complex bitstream analysis.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If deep bitstream analysis is performed to identify independently decodable pictures, then decoding accuracy is improved, but computational resources required worsen

Engineering Contradiction:
Improvedecoding accuracyVSAvoidcomputational resources
Core Design Contradiction:
Measurement precisionVSUse of energy by moving object

Solution Approach 1:

Picture independence information is prepared in advance by the encoder and embedded in SEI messages. This preliminary preparation shifts the computational burden to the encoding stage, allowing the decoder to simply read and act on the pre-computed independence declarations without performing complex analysis.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The SEI message serves as an intermediary data structure that efficiently encapsulates picture independence information. This standardized format reduces the computational complexity of identifying independent pictures compared to analyzing the full bitstream structure.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Productivity

If SEI message with picture independence indication is used, then parallel decoding capability is improved, but device complexity worsens

Engineering Contradiction:
Improveparallel decoding capabilityVSAvoiddecoder complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The decoding process is segmented into independent picture decoding tasks based on SEI message indications. Each independent picture can be assigned to a separate decoder instance, dividing the overall decoding workload into manageable, non-dependent segments that can be processed simultaneously.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The SEI message structure is designed to be universally applicable across different picture types and coding configurations. The same SEI message mechanism works for various picture formats and dependency patterns, reducing the need for specialized handling logic in the decoder.

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

Data Source

PatentUS20250119567A1SEI message supporting decoder picture-based parallelization
Publication Date: 2025.04.10 TENCENT AMERICA LLC
  • US20250119567A1 patent drawing
  • US20250119567A1 patent drawing
  • US20250119567A1 patent drawing

AI summary

A method and apparatus comprising computer code configured to cause a processor or processors to obtain video data including at least a first coded picture and a second coded picture, identify, by a decoder, at least one syntax element in a supplementary information enhancement (SEI) message indicating that the second coded picture is to be decoded independently from the first picture, and decode the second coded picture independently from the first coded picture based on the at least one syntax element in the SEI message.