TY - CONF
T1 - Peernets
T2 - 7th International Conference on Learning Representations, ICLR 2019
AU - Svoboda, Jan
AU - Masci, Jonathan
AU - Monti, Federico
AU - Bronstein, Michael M.
AU - Guibas, Leonidas
N1 - Publisher Copyright:
© 7th International Conference on Learning Representations, ICLR 2019. All Rights Reserved.
PY - 2019
Y1 - 2019
N2 - Deep learning systems have become ubiquitous in many aspects of our lives. Unfortunately, it has been shown that such systems are vulnerable to adversarial attacks, making them prone to potential unlawful uses. Designing deep neural networks that are robust to adversarial attacks is a fundamental step in making such systems safer and deployable in a broader variety of applications (e.g. autonomous driving), but more importantly is a necessary step to design novel and more advanced architectures built on new computational paradigms rather than marginally building on the existing ones. In this paper we introduce PeerNets, a novel family of convolutional networks alternating classical Euclidean convolutions with graph convolutions to harness information from a graph of peer samples. This results in a form of non-local forward propagation in the model, where latent features are conditioned on the global structure induced by the graph, that is up to 3× more robust to a variety of white- and black-box adversarial attacks compared to conventional architectures with almost no drop in accuracy.
AB - Deep learning systems have become ubiquitous in many aspects of our lives. Unfortunately, it has been shown that such systems are vulnerable to adversarial attacks, making them prone to potential unlawful uses. Designing deep neural networks that are robust to adversarial attacks is a fundamental step in making such systems safer and deployable in a broader variety of applications (e.g. autonomous driving), but more importantly is a necessary step to design novel and more advanced architectures built on new computational paradigms rather than marginally building on the existing ones. In this paper we introduce PeerNets, a novel family of convolutional networks alternating classical Euclidean convolutions with graph convolutions to harness information from a graph of peer samples. This results in a form of non-local forward propagation in the model, where latent features are conditioned on the global structure induced by the graph, that is up to 3× more robust to a variety of white- and black-box adversarial attacks compared to conventional architectures with almost no drop in accuracy.
UR - http://www.scopus.com/inward/record.url?scp=85083950059&partnerID=8YFLogxK
M3 - ???researchoutput.researchoutputtypes.contributiontoconference.paper???
AN - SCOPUS:85083950059
Y2 - 6 May 2019 through 9 May 2019
ER -