Inter-layer Prediction Using Temporal Sub-layer Information

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing video encoding and decoding technologies face challenges in efficiently supporting scalability and quality across diverse terminals and networks, particularly in handling high-definition and ultra-high-definition image compression and processing, due to varying terminal capabilities and network conditions.

Innovation Solution

An inter-layer prediction method using temporal sub-layer information is introduced, where the same maximum temporal sub-layer information is applied to all layers, reducing signaling overhead by omitting separate transmission of relation information and utilizing an indicator to control inter-layer prediction.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If separate maximum temporal sub-layer information is transmitted for each layer, then inter-layer prediction accuracy is improved, but signaling overhead increases

Engineering Contradiction:
Improveinter-layer prediction accuracyVSAvoidsignaling overhead
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent merges the maximum temporal sub-layer information into a single common value that is applied across all layers. Instead of transmitting separate maximum temporal sub-layer information for each layer, the system determines one maximum temporal sub-layer value that is valid for all layers, thereby reducing signaling overhead while maintaining prediction functionality.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent makes the maximum temporal sub-layer information universal across all layers by applying the same value to every layer. This single piece of information serves multiple layers simultaneously, eliminating the need for layer-specific parameters and reducing overall signaling requirements.

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

2Measurement precision

If detailed temporal sub-layer information is signaled for each layer, then prediction control precision is improved, but device complexity increases

Engineering Contradiction:
Improveprediction control precisionVSAvoidsignaling structure complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts the maximum temporal sub-layer information from the layer-specific context and places it in a common, universal context. By taking out the layer-specific maximum temporal sub-layer parameters and replacing them with a single common parameter, the system reduces signaling structure complexity while retaining essential prediction control.

Inventive Principle:
Principle #2Taking out (Extraction)

3Adaptability or versatility

If layer-specific maximum temporal sub-layer information is transmitted, then adaptability to different layer requirements is improved, but information transmission efficiency deteriorates

Engineering Contradiction:
Improvelayer-specific adaptationVSAvoidinformation transmission efficiency
Core Design Contradiction:
Adaptability or versatilityVSProductivity

Solution Approach 1:

The patent changes the parameter structure from multiple layer-specific maximum temporal sub-layer values to a single common maximum temporal sub-layer value. This parameter change simplifies the information transmission while maintaining the ability to control inter-layer prediction across all layers uniformly.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3913920A1Method and apparatus for predicting inter-layer based on temporal sub-layer information
Publication Date: 2021.11.24 ELECTRONICS & TELECOMM RES INST
  • EP3913920A1 patent drawingFigure 1
  • EP3913920A1 patent drawingFigure 2
  • EP3913920A1 patent drawingFigure 3

AI summary

The present invention relates to a method for predicting an inter-layer of an image having a plurality of layers including at least one temporal sub-layer. The method according to present invention comprises the steps of: acquiring information on a temporal sub-layer for inter-layer prediction; inducing a reference picture used for predicting an inter-layer of a current picture on the basis of the information on the temporal sub-layer; and predicting the inter-layer of the current picture on the basis of the reference picture.