DCT2 Enabled Flag for Video Coding Transform Selection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing video coding technologies do not allow for the exclusive use of the Discrete Cosine Transform type 2 (DCT2) transform core for all cases, limiting flexibility and efficiency in video compression.

Innovation Solution

The introduction of a sequence level DCT2 enabled flag allows for the inference of whether only DCT2 should be used for all cases, enabling a switch to pure DCT2 transform core when the flag is true, and allowing for Multiple Transform Selection (MTS) when the flag is false.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If Multiple Transform Selection (MTS) is enabled at sequence level, then transform flexibility and coding efficiency are improved, but the ability to use simple DCT2 transform core is lost

Engineering Contradiction:
Improvetransform flexibilityVSAvoidtransform core complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent introduces a dynamic flag `sps_dct2_enabled_flag` that allows the transform core selection to be adjusted based on coding conditions. When this flag is true, only DCT2 transform is used; when false, MTS can be enabled for greater flexibility. This dynamic control mechanism resolves the contradiction by allowing the system to adapt between simplicity and versatility depending on the specific coding requirements.

Inventive Principle:
Principle #15Dynamics

2Productivity

If DCT2 transform core is used exclusively, then computational complexity and memory bandwidth usage are reduced, but transform selection flexibility is limited

Engineering Contradiction:
Improvecomputational efficiencyVSAvoidtransform selection flexibility
Core Design Contradiction:
ProductivityVSAdaptability or versatility

Solution Approach 1:

The patent changes the parameter space by introducing the `sps_dct2_enabled_flag` as a control parameter. This flag allows the system to switch between two distinct transform processing modes: DCT2-only mode (when flag is true) for maximum efficiency, and MTS mode (when flag is false) for maximum flexibility. This parameter change enables the system to optimize between computational efficiency and transform selection flexibility based on coding conditions.

Inventive Principle:
Principle #35Parameter changes

3Productivity

If sequence level MTS is enabled, then coding efficiency is improved, but consistency in transform application across all cases is compromised

Engineering Contradiction:
Improvecoding efficiencyVSAvoidtransform application consistency
Core Design Contradiction:
ProductivityVSStability of the object's composition

Solution Approach 1:

The patent segments the transform application logic into two distinct paths controlled by the `sps_dct2_enabled_flag`. This segmentation ensures that when DCT2 is enabled, a consistent and simplified transform application is applied across all cases, while when MTS is enabled, flexibility is provided for specific blocks. This segmentation resolves the contradiction by allowing consistency where needed while maintaining efficiency where possible.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentUS12301807B2Encoder, decoder and corresponding methods using DCT2 enabled high level flag
Publication Date: 2025.05.13 HUAWEI TECH CO LTD
  • US12301807B2 patent drawing
  • US12301807B2 patent drawing
  • US12301807B2 patent drawing

AI summary

A method of video coding a block of a picture includes: obtaining a residual signal resulting from inter- or intra-picture prediction; inferring use of a DCT2 transform core for a sequence of the residual signal; processing the transformation of the block using the inferred transform core.