Timezone: »

 
Poster
Invariance-Aware Randomized Smoothing Certificates
Jan Schuchardt · Stephan Günnemann

Wed Nov 30 02:00 PM -- 04:00 PM (PST) @ Hall J #417
in Poster Session 4 »

Building models that comply with the invariances inherent to different domains, such as invariance under translation or rotation, is a key aspect of applying machine learning to real world problems like molecular property prediction, medical imaging, protein folding or LiDAR classification. For the first time, we study how the invariances of a model can be leveraged to provably guarantee the robustness of its predictions. We propose a gray-box approach, enhancing the powerful black-box randomized smoothing technique with white-box knowledge about invariances. First, we develop gray-box certificates based on group orbits, which can be applied to arbitrary models with invariance under permutation and Euclidean isometries. Then, we derive provably tight gray-box certificates. We experimentally demonstrate that the provably tight certificates can offer much stronger guarantees, but that in practical scenarios the orbit-based method is a good approximation.

Keywords: randomized smoothing · Invariances · Robustness certification · Verification · Equivariances
[ Paper]  [ Poster]  [ OpenReview] 

Author Information

Jan Schuchardt (Department of Informatics, Technical University Munich)
Stephan Günnemann (Technical University of Munich)

More from the Same Authors