Fault Injection Attacks on SoftMax Function in Deep Neural Networks Extended Abstract

被引：7

作者：

Jap, Dirmanto ^{[1
]}

Won, Yoo-Seung ^{[1
]}

Bhasin, Shivam ^{[1
]}

机构：

[1] Nanyang Technol Univ, Temasek Labs, Singapore, Singapore

来源：

PROCEEDINGS OF THE 18TH ACM INTERNATIONAL CONFERENCE ON COMPUTING FRONTIERS 2021 (CF 2021) | 2021年

关键词：

fault attacks; neural networks; softmax;

D O I：

10.1145/3457388.3458870

中图分类号：

TP301 [理论、方法];

学科分类号：

081202 ;

摘要：

Softmax is commonly used activation function in neural networks to normalize the output to probability distribution over predicted classes. Being often deployed in the output layer, it can potentially be targeted by fault injection attacks to create misclassification. In this extended abstract, we perform a preliminary fault analysis of Softmax against single bit faults.

引用

页码：238 / 240

页数：3

共 50 条

[1] Fault Injection Attacks in Spiking Neural Networks and Countermeasures
Nagarajan, Karthikeyan
Li, Junde
Ensan, Sina Sayyah
Kannan, Sachhidh
Ghosh, Swaroop
[J]. FRONTIERS IN NANOTECHNOLOGY, 2022, 3
[2] HASHTAG: Hash Signatures for Online Detection of Fault-Injection Attacks on Deep Neural Networks
Javaheripi, Mojan
Koushanfar, Farinaz
[J]. 2021 IEEE/ACM INTERNATIONAL CONFERENCE ON COMPUTER AIDED DESIGN (ICCAD), 2021,
[3] Deep Learning of Graphs with Ngram Convolutional Neural Networks(Extended abstract)
Luo, Zhiling
Liu, Ling
Yin, Jianwei
Li, Ying
Wu, Zhaohui
[J]. 2018 IEEE 34TH INTERNATIONAL CONFERENCE ON DATA ENGINEERING (ICDE), 2018, : 1791 - 1792
[4] Efficient Softmax Hardware Architecture for Deep Neural Networks
Du, Gaoming
Tian, Chao
Li, Zhenmin
Zhang, Duoli
Yin, Yongsheng
Ouyang, Yiming
[J]. GLSVLSI '19 - PROCEEDINGS OF THE 2019 ON GREAT LAKES SYMPOSIUM ON VLSI, 2019, : 75 - 80
[5] Robust Adversarial Attacks on Imperfect Deep Neural Networks in Fault Classification
Jiang, Xiaoyu
Kong, Xiangyin
Zheng, Junhua
Ge, Zhiqiang
Zhang, Xinmin
Song, Zhihuan
[J]. IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, 2024,
[6] Causality in Neural Networks - An Extended Abstract
Reddy, Abbavaram Gowtham
[J]. AIES '21: PROCEEDINGS OF THE 2021 AAAI/ACM CONFERENCE ON AI, ETHICS, AND SOCIETY, 2021, : 271 - 272
[7] Analysis of Power-Oriented Fault Injection Attacks on Spiking Neural Networks
Nagarajan, Karthikeyan
Li, Junde
Ensan, Sina Sayyah
Khan, Mohammad Nasim Imtiaz
Kannan, Sachhidh
Ghosh, Swaroop
[J]. PROCEEDINGS OF THE 2022 DESIGN, AUTOMATION & TEST IN EUROPE CONFERENCE & EXHIBITION (DATE 2022), 2022, : 861 - 866
[8] On the Learning Property of Logistic and Softmax Losses for Deep Neural Networks
Li, Xiangrui
Li, Xin
Pan, Deng
Zhu, Dongxiao
[J]. THIRTY-FOURTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, THE THIRTY-SECOND INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE AND THE TENTH AAAI SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2020, 34 : 4739 - 4746
[9] enpheeph: A Fault Injection Framework for Spiking and Compressed Deep Neural Networks
Colucci, Alessio
Steininger, Andreas
Shafique, Muhammad
[J]. 2022 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS), 2022, : 5155 - 5162
[10] Hardware-Aware Softmax Approximation for Deep Neural Networks
Geng, Xue
Lin, Jie
Zhao, Bin
Kong, Anmin
Aly, Mohamed M. Sabry
Chandrasekhar, Vijay
[J]. COMPUTER VISION - ACCV 2018, PT IV, 2019, 11364 : 107 - 122

← 1 2 3 4 5 →