University of Twente Student Theses
Learning data augmentation policies for computer vision using additive Fourier-basis noise
Zeng, Y. (2023) Learning data augmentation policies for computer vision using additive Fourier-basis noise.
|
PDF
6MB |
| Abstract: | Data augmentation is an important tool to improve model robustness. This study uses Fourier-basis noise to augment images. A new approach is introduced that utilizes Reinforcement Learning to find useful combinations of noise as augmentation policies. The results demonstrate that the searched Fourier-basis augmentation is more effective in improving the model's robustness to corruption than the baseline model. Furthermore, combining different augmentation techniques further enhances the model's performance, indicating that Fourier-basis augmentation positively affects model robustness. |
| Item Type: | Essay (Master) |
| Faculty: | EEMCS: Electrical Engineering, Mathematics and Computer Science |
| Subject: | 54 computer science |
| Programme: | Computer Science MSc (60300) |
| Link to this item: | https://purl.utwente.nl/essays/94515 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page