Publication | Closed Access
Fast and Effective Robustness Certification
260
Citations
0
References
2018
Year
EngineeringVerificationRobustness TestingSoftware AnalysisFormal VerificationHardware SecurityEffective Robustness CertificationSparse Neural NetworkAdversarial Machine LearningSystems EngineeringNeural Network RobustnessReliabilitySoftware CertificationAbstract InterpretationComputer EngineeringComputer ScienceNeural NetworksDeep LearningNeural Architecture SearchData SecurityAttack ModelSoftware Testing
We present a new method and system, called DeepZ, for certifying neural network robustness based on abstract interpretation. Compared to state-of-the-art automated verifiers for neural networks, DeepZ: (i) handles ReLU, Tanh and Sigmoid activation functions, (ii) supports feedforward and convolutional architectures, (iii) is significantly more scalable and precise, and (iv) and is sound with respect to floating point arithmetic. These benefits are due to carefully designed approximations tailored to the setting of neural networks. As an example, DeepZ achieves a verification accuracy of 97% on a large network with 88,500 hidden units under $L_{\infty}$ attack with $\epsilon = 0.1$ with an average runtime of 133 seconds.