Web: http://arxiv.org/abs/2103.03076

March 15, 2023, 1:10 a.m. | Fu Wang, Yanghao Zhang, Yanbin Zheng, Wenjie Ruan

cs.CR updates on arXiv.org arxiv.org

Adversarial training is an effective but time-consuming way to train robust
deep neural networks that can withstand strong adversarial attacks. As a
response to its inefficiency, we propose Dynamic Efficient Adversarial Training
(DEAT), which gradually increases the adversarial iteration during training. We
demonstrate that the gradient's magnitude correlates with the curvature of the
trained model's loss landscape, allowing it to reflect the effect of
adversarial training. Therefore, based on the magnitude of the gradient, we
propose a general acceleration strategy, …

adversarial dynamic magnitude training

OCIO-0020 Expert on Enterprise Cyber-related Exercises Support (NS) - MON 3 Apr

@ EMW, Inc. | Brussels, Brussels, Belgium

Senior Consultant (m/w/d) IT Security Management - remote/ onsite

@ MVI Group GmbH | München, Germany

Cyber Security Analyst II - Remote Contract (3913-W)

@ Stout Systems | New York City, United States - Remote

Sr Cyber Security Advisor - S Atlantic Enterprise

@ Optiv | Atlanta, GA

Director, Information Security Operations (Ottawa, ON)

@ SSENSE | Ottawa, ON, Canada

Senior Security Engineer - Threat Detection

@ Samsara | San Francisco, CA

Information Security Engineer

@ NMI | Bristol, England, United Kingdom - Remote

Co-op Student - IT Compliance

@ BusPatrol | Remote

Incident Response Manager

@ matchpoint solutions | Houston, Tx, Houston, Tx, United States

Consultant(e) réseau / sécurité H/F - Innovative Tech

@ Devoteam | Nantes, France

DevSecOps Security Consultant 2023 ECL

@ Bosch Group | Bengaluru, India

Staff Back-end Engineer [Security Architecture & Engineering]

@ Coupang | Seoul, South Korea