Understanding non-robust features in image classification
Name
1193031569-MIT.pdf
Size
7.39 MB
Format
Adobe PDF
Checksum (MD5)
93a48fad4b36af060f28001ac0bbca24
Author(s)
Wei, Kuo-An Andy.
Advisor(s)
Aleksander Mądry.
Date Issued
2020
Publisher
Massachusetts Institute of Technology
Abstract
Despite the remarkable success of deep neural networks on image classification tasks, they exhibit a surprising vulnerability to certain small worst-case perturbations, also known as adversarial examples. Over the years, many different theories have been proposed to explain this puzzling phenomenon. Recent work by Ilyas et al. proposes a fresh new take on the existence of adversarial examples--that adversarial examples are inevitable due to certain well-generalizing but non-robust features present in the natural data [14]. We build upon the "non-robust features" framework raised by Ilyas et al., and present some new observations on the properties of non-robust features. We showcase some visualization techniques based on adversarial attacks to help us build an intuitive understanding of non-robust features. Lastly, we propose a novel framework for analyzing the types of information present in non-robust features, known as the adversarial transferability analysis.
Description
Thesis: M. Eng., Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, May, 2020
Cataloged from the official PDF of thesis.
Includes bibliographical references (pages 39-41).
Subjects
Electrical Engineering and Computer Science.
MIT Department
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Terms of Use
MIT theses may be protected by copyright. Please reuse MIT thesis content according to the MIT Libraries Permissions Policy, which is available through the URL provided.
Persistent DSpace Link
