FILTER PRUNING VIA SOFTMAX ATTENTION
- Authors
- Cho, S.; Kim, H.; Kwon, Junseok
- Issue Date
- Aug-2021
- Publisher
- IEEE Computer Society
- Keywords
- Relative depth-wise separable convolutions; Softmax attention channel pruning
- Citation
- Proceedings - International Conference on Image Processing, ICIP, v.2021-September, pp 3507 - 3511
- Pages
- 5
- Journal Title
- Proceedings - International Conference on Image Processing, ICIP
- Volume
- 2021-September
- Start Page
- 3507
- End Page
- 3511
- URI
- https://scholarworks.bwise.kr/cau/handle/2019.sw.cau/55494
- DOI
- 10.1109/ICIP42928.2021.9506724
- ISSN
- 1522-4880
- Abstract
- In this paper, we propose a novel network pruning method using the proposed relative depth-wise separable convolutions and softmax attention channel pruning. The relative depth-wise separable convolution enhances conventional depth-wise separable convolutions by enabling the channel interaction, which can prevent accuracy drops even after severe pruning. The softmax attention channel pruning probabilistically expresses the importance of filters and removes unimportant channels efficiently. Experimental results demonstrate that our pruning method outperforms other state-of-the-art pruning methods in terms of Flops, parameters, and top-1 classification accuracy. © 2021 IEEE.
- Files in This Item
- There are no files associated with this item.
- Appears in
Collections - College of Software > School of Computer Science and Engineering > 1. Journal Articles
![qrcode](https://api.qrserver.com/v1/create-qr-code/?size=55x55&data=https://scholarworks.bwise.kr/cau/handle/2019.sw.cau/55494)
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.