BACHELOR Assignment
Fourier-basis noise data augmentation.
Type: Bachelor CS
Period: TBD
Student: (Unassigned)
If you are interested please contact :
Description:
Traditional image augmentations include but are not limited to rotation, crop, and flip, which mainly operate in the spatial domain. By using image data augmentation, the variety of the data increases, improving model robustness. Instead of operating in the spatial domain, this research aims at exploring whether augmenting the frequency information of images helps improve the robustness of models under adversarial attacks [2]. To augment frequency information, Fourier basis noise is used [1], which is additive noise in a specific frequency. The frequency-augmented images are further used for training computer vision models, for which the robustness towards adversarial attack is evaluated and compared with state-of-the-art approaches.
References:
[1] Yin, D., Lopes, R. G., Shlens, J., Cubuk, E. D., & Gilmer, J. (2019). A Fourier Perspective on Model Robustness in Computer Vision. doi:10.48550/ARXIV.1906.08988
[2] S. Y. Khamaiseh, D. Bagagem, A. Al-Alaj, M. Mancino and H. W. Alomari, "Adversarial Deep Learning: A Survey on Adversarial Attacks and Defense Mechanisms on Image Classification," in IEEE Access, vol. 10, pp. 102266-102291, 2022, doi: 10.1109/ACCESS.2022.3208131