Compound Orthonormal Transform for Video Residual Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding standards, such as HEVC, face inefficiencies in transform core implementations, particularly with the discrete sine transform (DST)-7 and discrete cosine transform (DCT)-8, which have higher complexity and memory costs compared to DST-4 and DCT-4, and require additional storage for transform bases, while also showing better coding performance.
Innovation Solution
Embedding N-point DST-7/DCT-8 into 2N-point DCT-2 transforms to create a Compound Orthonormal Transform (COT), sharing logic and reducing memory costs by utilizing symmetric/anti-symmetric characteristics, and applying these transforms selectively based on block sizes for improved coding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If DST-7 and DCT-8 transforms are used for video coding, then coding performance is improved, but transform complexity and memory cost increase
Solution Approach 1:
The patent combines DST-7 and DCT-8 transforms into a unified Compound Orthonormal Transform (COT) framework. By merging these transforms and exploiting their shared symmetric/anti-symmetric characteristics, the implementation reduces computational complexity while maintaining the coding performance benefits of both transforms. The COT allows selective application of DST-7 or DCT-8 based on block characteristics without requiring separate full implementations of both transforms.
Solution Approach 2:
The COT structure serves multiple functions: it can perform both DST-7 and DCT-8 transforms using a single unified framework, supports selective application based on block size and characteristics, and provides a common implementation path that reduces overall system complexity. This multi-functional approach eliminates the need for separate dedicated implementations of DST-7 and DCT-8.
2Manufacturing precision
If DST-7 and DCT-8 transforms are used for video coding, then coding performance is improved, but memory cost increases
Solution Approach 1:
By merging DST-7 and DCT-8 into the COT framework, the patent eliminates redundant storage of transform bases. The unified structure allows both transforms to share common computational elements and lookup tables, significantly reducing the memory required to store transform coefficients and intermediate results compared to maintaining separate full implementations.
Solution Approach 2:
The COT implementation uses efficient copying and reuse of computational patterns. Symmetric and anti-symmetric properties are exploited to copy and reuse transform basis functions across different transform types, reducing the need to store complete sets of basis functions for both DST-7 and DCT-8 separately.
3Productivity
If selective transform application is implemented, then coding efficiency is optimized, but decision complexity increases
Solution Approach 1:
The patent applies different transforms (DST-7 or DCT-8) based on local characteristics of video blocks, such as block size and residual energy distribution. This local quality approach allows the encoder to select the most appropriate transform for each specific block, optimizing coding efficiency while using simple, localized decision criteria rather than complex global optimization.
Solution Approach 2:
The selection between DST-7 and DCT-8 is based on changing key parameters of the video block, such as block size, transform coefficient energy distribution, and prediction mode. These parameter-based selection criteria provide a straightforward decision mechanism that balances coding efficiency optimization with minimal decision complexity.
Data Source
AI summary
A method of controlling residual coding for decoding or encoding of a video sequence, is performed by at least one processor and includes determining whether a small transform size of a primary transform is to be used for the residual coding of a coded block of the video sequence. The method further includes based on the small transform size of the primary transform being determined to be used, identifying, as the primary transform, a first transform set including discrete sine transform (DST)-4 and discrete cosine transform (DCT)-4, based on the small transform size of the primary transform being determined to not be used, identifying, as the primary transform, a second transform set including DST-7 and DCT-8, and performing the residual coding of the coded block, using the identified primary transform.