Compressing image-to-image models with average smoothing
US11790565B2 · kind B2 · utility
Assignee
Inventors
Key dates
| Filing date | Mar 4, 2021 |
| Grant date | Oct 17, 2023 |
| Priority date | — |
| Expiry date | Dec 16, 2041 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG06N3/0985
- WIPO fieldAudio-visual technology
- WIPO sectorElectrical engineering
Abstract
System and methods for compressing image-to-image models. Generative Adversarial Networks (GANs) have achieved success in generating high-fidelity images. An image compression system and method adds a novel variant to class-dependent parameters (CLADE), referred to as CLADE-Avg, which recovers the image quality without introducing extra computational cost. An extra layer of average smoothing is performed between the parameter and normalization layers. Compared to CLADE, this image compression system and method smooths abrupt boundaries, and introduces more possible values for the scaling and shift. In addition, the kernel size for the average smoothing can be selected as a hyperparameter, such as a 3×3 kernel size. This method does not introduce extra multiplications but only addition, and thus does not introduce much computational overhead, as the division can be absorbed into the parameters after training.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.