Technical Program

AML-L.1: Adversarial Machine Learning I

Symposium: Signal Processing for Adversarial Machine Learning
Session Type: Lecture
Time: Thursday, November 29, 14:00 - 15:30
Location: Castle
Session Chairs: Sijia Liu, IBM and Pin-Yu Chen, IBM Research AI
 
14:00 - 14:18
AML-L.1.1: DIFFERENTIALLY PRIVATE SPARSE INVERSE COVARIANCE ESTIMATION
         Di Wang; State University of New York at Buffalo
         Mengdi Huai; State University of New York at Buffalo
         Jinhui Xu; State University of New York at Buffalo
 
14:18 - 14:36
AML-L.1.2: DEFENDING DNN ADVERSARIAL ATTACKS WITH PRUNING AND LOGITS AUGMENTATION
         Siyue Wang; Northeastern University
         Xiao Wang; Boston University
         Shaokai Ye; Syracuse University
         Pu Zhao; Northeastern University
         Xue Lin; Northeastern University
 
14:36 - 14:54
AML-L.1.3: ON THE UTILITY OF CONDITIONAL GENERATION BASED MUTUAL INFORMATION FOR CHARACTERIZING ADVERSARIAL SUBSPACES
         Chia-Yi Hsu; National Chung Hsing University
         Pei-Hsuan Lu; National Chung Hsing University
         Pin-Yu Chen; IBM Research AI
         Chia-Mu Yu; National Chung Hsing University
 
14:54 - 15:12
AML-L.1.4: BACKDOOR ATTACKS ON NEURAL NETWORK OPERATIONS
         Joseph Clements; Clemson University
         Yingjie Lao; Clemson University
 
15:12 - 15:30
AML-L.1.5: ON EXTENSIONS OF CLEVER: A NEURAL NETWORK ROBUSTNESS EVALUATION ALGORITHM
         Tsui-Wei Weng; Massachusetts Institute of Technology
         Huan Zhang; University of California, Davis
         Pin-Yu Chen; IBM Research AI
         Aurelie Lozano; IBM Research AI
         Cho-Jui Hsieh; University of California, Davis
         Luca Daniel; Massachusetts Institute of Technology