Regularizing Activation Distribution for Ultra Low-bit Quantization-Aware Training of MobileNets
- Authors
- Park, Seongmin; Sung, Wonyong; Choi, Jungwook
- Issue Date
- Nov-2022
- Publisher
- Institute of Electrical and Electronics Engineers Inc.
- Keywords
- CNN; MobileNet; Quantization-aware training
- Citation
- IEEE Workshop on Signal Processing Systems, SiPS: Design and Implementation, v.2022-November, pp.1 - 6
- Indexed
- SCOPUS
- Journal Title
- IEEE Workshop on Signal Processing Systems, SiPS: Design and Implementation
- Volume
- 2022-November
- Start Page
- 1
- End Page
- 6
- URI
- https://scholarworks.bwise.kr/hanyang/handle/2021.sw.hanyang/172850
- DOI
- 10.1109/SiPS55645.2022.9919240
- ISSN
- 1520-6130
- Abstract
- MobileNets are a family of CNN architectures that are designed for parameter efficient model deployment. However, quantization aware training (QAT) of MobileNets for lowprecision model development has not been very successful partly due to the parameter efficient nature of the model. The activation signal in MobileNet shows fairly large magnitude diversity for each channel, which can lead to poor quantization results. We address this issue by including a new loss term for QAT, which regularizes the diversity of the activation signal in each channel. The proposed method improves the accuracy of the state-of-theart QAT when evaluated on MobileNet with CIFAR10/100, and ImageNet datasets. © 2022 IEEE.
- Files in This Item
-
Go to Link
- Appears in
Collections - 서울 공과대학 > 서울 융합전자공학부 > 1. Journal Articles
![qrcode](https://api.qrserver.com/v1/create-qr-code/?size=55x55&data=https://scholarworks.bwise.kr/hanyang/handle/2021.sw.hanyang/172850)
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.