2 papers
cs.LG2020
Synthesizing Unrestricted False Positive Adversarial Objects Using Generative Models
Martin Kotuliak, Sandro E. Schoenborn, Andrei Dan
Adversarial examples are data points misclassified by neural networks. Originally, adversarial examples were limited to adding small perturbations to a given image. Recent work int…
cs.LG2020
Scalable Polyhedral Verification of Recurrent Neural Networks
Wonryong Ryou, Jiayu Chen, Mislav Balunovic +3
We present a scalable and precise verifier for recurrent neural networks, called Prover based on two novel ideas: (i) a method to compute a set of polyhedral abstractions for the n…