Media Track Encapsulation for Selective Temporal Layer Decoding

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional video coding technologies, such as AVS3, only support single-track encapsulation of video bitstreams containing multiple temporal layers, making it difficult for users to selectively reconstruct the bitstream via different combinations of tracks, which limits flexibility and computation resource utilization.

Innovation Solution

The method encapsulates media frames of multiple temporal layers into different tracks, with each track containing specific temporal layer information, and marks a base track, allowing content consuming devices to select appropriate layers for decoding and displaying, thereby ensuring flexibility and optimizing computation resources.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If video bitstreams containing multiple temporal layers are encapsulated as a single track, then the encapsulation structure is simple, but users cannot selectively reconstruct the bitstream via different combinations of tracks, limiting flexibility

Engineering Contradiction:
Improveflexibility in video reconstructionVSAvoidencapsulation structure complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent segments the video bitstream by creating separate tracks for different temporal layers (e.g., Track 1 for temporal layer 0, Track 2 for temporal layer 1, etc.). Each track contains media frames belonging to a specific temporal layer, allowing independent selection and combination of tracks based on playback requirements. This segmentation enables flexible reconstruction while maintaining clear organizational structure through track-level separation.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces a new organizational dimension by mapping temporal layers to separate tracks. Instead of organizing video data in a single-dimensional track structure, it creates a multi-dimensional organization where temporal layer information becomes a separate dimension (track identifier). This allows content consuming devices to select specific temporal layers by choosing appropriate tracks, thereby achieving flexibility without excessive complexity.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Productivity

If all temporal layers are included in a single track, then the track structure is simple, but computation resources are not optimized as devices cannot select only needed layers

Engineering Contradiction:
Improvecomputation resource utilization efficiencyVSAvoidmulti-track management complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent extracts temporal layer information from a unified track structure and places each temporal layer into its own dedicated track. This extraction allows content consuming devices to selectively decode only the temporal layers they need (e.g., only layer 0 for basic playback, or layers 0 and 1 for enhanced quality), thereby optimizing computation resource utilization by avoiding unnecessary decoding of higher temporal layers when they are not required.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent enables partial action by allowing devices to decode only the necessary portion of the video content (specific temporal layers) rather than requiring full decoding of all layers. Devices can perform partial decoding based on their capabilities and requirements, improving computation resource efficiency while the multi-track structure manages the complexity of having multiple temporal layers available.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP4354868B1Media data processing method and related device
Publication Date: 2025.12.10 TENCENT TECHNOLOGY (SHENZHEN) CO LTD
  • EP4354868B1 patent drawingFigure 1~2
  • EP4354868B1 patent drawingFigure 3~5
  • EP4354868B1 patent drawingFigure 6a~7

AI summary

Provided in the embodiments of the present application are a media data processing method and a related device. The method comprises: acquiring a description data box of a jth track among M tracks, wherein the description data box comprises time-domain track indication information, the time-domain track indication information is used for indicating track encapsulation manners for N time-domain levels, and comprises time-domain level information of a time-domain level encapsulated in the jth track, j is a positive integer, and j ≤ M; and decoding media data according to the time-domain track indication information. The flexibility of performing multi-track encapsulation on different time-domain levels can thus be improved.