PPT-Robustness to Adversarial Examples

Presenters Pooja Harekoppa Daniel Friedman Explaining and Harnessing Adversarial Examples Ian J Goodfellow Jonathon Shlens and Christian Szegedy Google Inc Mountain View CA Highlights . in . Encryption Schemes. Payman. . Mohassel. University of Calgary. Public Key Encryption (PKE). pk. (. pk. , . sk. )  . KG. C = Enc(. pk,m. ). m = Dec(. sk,C. ) . PKE = (KG, Enc, Dec). 2. Traditional Security Notions. Aditya. . Parameswaran. Stanford University. (Joint work with: . Nilesh. . Dalvi. , Hector Garcia-Molina, . Rajeev . Rastogi. ). 1. 2. 3. html. body. head. title. div. div. table. td. table. etc. Convnets. (optimize weights to predict bus). bus. Convnets. (optimize input to predict ostrich). ostrich. Work on Adversarial examples by . Goodfellow. et al. , . Szegedy. et. al., etc.. Generative Adversarial Networks (GAN) [. Statistical Relational AI. Daniel Lowd. University of Oregon. Outline. Why do we need adversarial modeling?. Because of the dream of AI. Because of current reality. Because of possible dangers. Our initial approach and results. Testing. . by- . . . (Group 2-Batch F1). . Robustness Testing. A Type Of Black . B. ox Testing. Black. -. box . testing.  is a method . Nets. İlke Çuğu 1881739. NIPS 2014 . Ian. . Goodfellow. et al.. At a . glance. (. http://www.kdnuggets.com/2017/01/generative-adversarial-networks-hot-topic-machine-learning.html. ). Idea. . Behind. Adversarial examples. Ostrich!. Adversarial examples. Ostrich!. Intriguing properties of neural networks. . Christian . Szegedy. , . Wojciech. . Zaremba. , Ilya . Sutskever. , Joan Bruna, . Dumitru. ML Reading . Group. Xiao Lin. Jul. 22 2015. I. . Goodfellow. , J. . Pouget-Abadie. , M. Mirza, B. Xu, D. . Warde. -Farley, S. . Ozair. , A. . Courville. and Y. . Bengio. . . "Generative adversarial nets." . Deep Learning and Security Workshop 2017. Chang Liu. UC Berkeley. Deep Learning and Security is a trending topic in academia in 2017. Best Papers in Security Conferences. Towards Evaluating the Robustness of Neural Networks (Oakland 2017 Best Student Paper). Paul E. Griffiths. 19/6/18. Robustness workshop, Bordeaux. 1. Developmental systems. “. An animal is, in fact, a developmental system, and it is these systems, not the mere adult forms which we conventionally take as typical of the species, which becomes modified during the course of evolution.”. Florian Tramèr. Stanford University, Google, ETHZ. ML suffers from . adversarial. . examples.. 2. 90% Tabby Cat. 100% Guacamole. Adversarial noise. Robust classification is . hard! . 3. Clean. Adversarial (. Dr. Alex Vakanski. Lecture 6. GANs for Adversarial Machine Learning. Lecture Outline. Mohamed Hassan presentation. Introduction to Generative Adversarial Networks (GANs). Jeffrey Wyrick presentation. Attacks. Haotian Wang. Ph.D. . . Student. University of Idaho. Computer Science. Outline. Introduction. Defense . a. gainst . Adversarial Attack Methods. Gradient Masking/Obfuscation. Robust Optimization. Dr. Alex Vakanski. Lecture 1. Introduction to Adversarial Machine Learning. . Lecture Outline. Machine Learning (ML). Adversarial ML (AML). Adversarial examples. Attack taxonomy. Common adversarial attacks.