Adversarial attacks and robust defenses in deep learning

Chun Pong Lau; Jiang Liu; Wei An Lin; Hossein Souri; Pirazh Khorramshahi; Rama Chellappa

doi:10.1016/bs.host.2023.01.001

Adversarial attacks and robust defenses in deep learning

Chun Pong Lau, Jiang Liu, Wei An Lin, Hossein Souri, Pirazh Khorramshahi, Rama Chellappa

Whiting School of Engineering

Research output: Chapter in Book/Report/Conference proceeding › Chapter

Abstract

Deep learning models have shown exceptional performance in many applications, including computer vision, natural language processing, and speech processing. However, if no defense strategy is considered, deep learning models are vulnerable to adversarial attacks. In this chapter, we will first describe various typical adversarial attacks. Then we will describe different adversarial defense methods for image classification and object detection tasks.

Original language	English (US)
Title of host publication	Deep Learning
Editors	Venu Govindaraju, Arni S.R. Srinivasa Rao, C.R. Rao
Publisher	Elsevier B.V.
Pages	29-58
Number of pages	30
ISBN (Print)	9780443184307
DOIs	https://doi.org/10.1016/bs.host.2023.01.001
State	Published - Jan 2023

Publication series

Name	Handbook of Statistics
Volume	48
ISSN (Print)	0169-7161

Keywords

Adversarial attacks
Deep learning
Defenses against adversarial attacks

ASJC Scopus subject areas

Statistics and Probability
Modeling and Simulation
Applied Mathematics

Access to Document

10.1016/bs.host.2023.01.001

Cite this

@inbook{41ef1371c71341599ffc977f7f08c4a0,

title = "Adversarial attacks and robust defenses in deep learning",

abstract = "Deep learning models have shown exceptional performance in many applications, including computer vision, natural language processing, and speech processing. However, if no defense strategy is considered, deep learning models are vulnerable to adversarial attacks. In this chapter, we will first describe various typical adversarial attacks. Then we will describe different adversarial defense methods for image classification and object detection tasks.",

keywords = "Adversarial attacks, Deep learning, Defenses against adversarial attacks",

author = "Lau, {Chun Pong} and Jiang Liu and Lin, {Wei An} and Hossein Souri and Pirazh Khorramshahi and Rama Chellappa",

note = "Funding Information: This work was supported by the DARPA GARD Program under the contract HR001119S0026-GARD-FP-052. Publisher Copyright: {\textcopyright} 2023 Elsevier B.V.",

year = "2023",

month = jan,

doi = "10.1016/bs.host.2023.01.001",

language = "English (US)",

isbn = "9780443184307",

series = "Handbook of Statistics",

publisher = "Elsevier B.V.",

pages = "29--58",

editor = "Venu Govindaraju and {Srinivasa Rao}, {Arni S.R.} and C.R. Rao",

booktitle = "Deep Learning",

}

TY - CHAP

T1 - Adversarial attacks and robust defenses in deep learning

AU - Lau, Chun Pong

AU - Liu, Jiang

AU - Lin, Wei An

AU - Souri, Hossein

AU - Khorramshahi, Pirazh

AU - Chellappa, Rama

PY - 2023/1

Y1 - 2023/1

N2 - Deep learning models have shown exceptional performance in many applications, including computer vision, natural language processing, and speech processing. However, if no defense strategy is considered, deep learning models are vulnerable to adversarial attacks. In this chapter, we will first describe various typical adversarial attacks. Then we will describe different adversarial defense methods for image classification and object detection tasks.

AB - Deep learning models have shown exceptional performance in many applications, including computer vision, natural language processing, and speech processing. However, if no defense strategy is considered, deep learning models are vulnerable to adversarial attacks. In this chapter, we will first describe various typical adversarial attacks. Then we will describe different adversarial defense methods for image classification and object detection tasks.

KW - Adversarial attacks

KW - Deep learning

KW - Defenses against adversarial attacks

UR - http://www.scopus.com/inward/record.url?scp=85148347387&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85148347387&partnerID=8YFLogxK

U2 - 10.1016/bs.host.2023.01.001

DO - 10.1016/bs.host.2023.01.001

M3 - Chapter

AN - SCOPUS:85148347387

SN - 9780443184307

T3 - Handbook of Statistics

SP - 29

EP - 58

BT - Deep Learning

A2 - Govindaraju, Venu

A2 - Srinivasa Rao, Arni S.R.

A2 - Rao, C.R.

PB - Elsevier B.V.

ER -

Adversarial attacks and robust defenses in deep learning

Abstract

Publication series

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this