Observations on K-image Expansion of Image-Mixing Augmentationopen access
- Authors
- Jeong, J.; Cha, S.; Choi, Jongwon; Yun, S.; Moon, T.; Yoo, Y.
- Issue Date
- 2023
- Publisher
- Institute of Electrical and Electronics Engineers Inc.
- Keywords
- Augmentation; Computer architecture; Data augmentation; Dirichlet process; Image classification; Image Classification; Measurement uncertainty; Probabilistic logic; Robustness; Uncertainty
- Citation
- IEEE Access, v.11, pp 1 - 1
- Pages
- 1
- Journal Title
- IEEE Access
- Volume
- 11
- Start Page
- 1
- End Page
- 1
- URI
- https://scholarworks.bwise.kr/cau/handle/2019.sw.cau/69552
- DOI
- 10.1109/ACCESS.2023.3243108
- ISSN
- 2169-3536
- Abstract
- Image-mixing augmentations (e.g., Mixup and CutMix), which typically involve mixing two images, have become the de-facto training techniques for image classification. Despite their huge success in image classification, the number of images to be mixed has not been elucidated in the literature: only the naive K-image expansion has been shown to lead to performance degradation. This study derives a new K-image mixing augmentation based on the stick-breaking process under Dirichlet prior distribution. We demonstrate superiority of our K-image expansion augmentation over conventional two-image mixing augmentation methods through extensive experiments and analyses: (1) more robust and generalized classifiers; (2) a more desirable loss landscape shape; (3) better adversarial robustness. Moreover, we show that our probabilistic model can measure the sample-wise uncertainty and boost the efficiency for network architecture search by achieving a 7-fold reduction in the search time. Author
- Files in This Item
-
- Appears in
Collections - Graduate School of Advanced Imaging Sciences, Multimedia and Film > Department of Imaging Science and Arts > 1. Journal Articles
![qrcode](https://api.qrserver.com/v1/create-qr-code/?size=55x55&data=https://scholarworks.bwise.kr/cau/handle/2019.sw.cau/69552)
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.