An abstract domain for certifying neural networks

Citations Over TimeTop 1% of 2019 papers

Abstract

We present a novel method for scalable and precise certification of deep neural networks. The key technical insight behind our approach is a new abstract domain which combines floating point polyhedra with intervals and is equipped with abstract transformers specifically tailored to the setting of neural networks. Concretely, we introduce new transformers for affine transforms, the rectified linear unit (ReLU), sigmoid, tanh, and maxpool functions. We implemented our method in a system called DeepPoly and evaluated it extensively on a range of datasets, neural architectures (including defended networks), and specifications. Our experimental results indicate that DeepPoly is more precise than prior work while scaling to large networks. We also show how to combine DeepPoly with a form of abstraction refinement based on trace partitioning. This enables us to prove, for the first time, the robustness of the network when the input image is subjected to complex perturbations such as rotations that employ linear interpolation.

Related Papers

• → Remarks on neural network controller using different sigmoid functions(2002)4 cited
• → Describing function approximation for biomedical engineering applications(2004)3 cited
• → New polyhedra – Part 3(2011)
• The nlogistic-sigmoid function(2020)
• → A Methodology for Increasing Precision for the Efficient Implementation of Sigmoid Function(2023)