Robustness Verification of Classification Deep Neural Networks via Linear Programming
Authors:
- 1 June 2019
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 11410-11419
Abstract
There is a pressing need to verify robustness of classification deep neural networks (CDNNs) as they are embedded in many safety-critical applications. Existing robustness verification approaches rely on computing the over-approximation of the output set, and can hardly scale up to practical CDNNs, as the result of error accumulation accompanied with approximation. In this paper, we develop a novel method for robustness verification of CDNNs with sigmoid activation functions. It converts the robustness verification problem into an equivalent problem of inspecting the most suspected point in the input region which constitutes a nonlinear optimization problem. To make it amenable, by relaxing the nonlinear constraints into the linear inclusions, it is further refined as a linear programming problem. We conduct comparison experiments on a few CDNNs trained for classifying images in some state-of-the-art benchmarks, showing our advantages of precision and scalability that enable effective verification of practical CDNNs.Keywords
This publication has 19 references indexed in Scilit:
- Adaptive Neural Network Control of Biped RobotsIEEE Transactions on Systems, Man, and Cybernetics: Systems, 2016
- A Combined Adaptive Neural Network and Nonlinear Model Predictive Control for Multirate Networked Industrial Process ControlIEEE Transactions on Neural Networks and Learning Systems, 2015
- Deep learning in neural networks: An overviewNeural Networks, 2015
- Evasion Attacks against Machine Learning at Test Time2013
- Deep Neural Networks for Acoustic Modeling in Speech Recognition: The Shared Views of Four Research GroupsIEEE Signal Processing Magazine, 2012
- Man vs. computer: Benchmarking machine learning algorithms for traffic sign recognitionNeural Networks, 2012
- An Abstraction-Refinement Approach to Verification of Artificial Neural Networks2010
- Learning generative visual models from few training examples: An incremental Bayesian approach tested on 101 object categoriesComputer Vision and Image Understanding, 2007
- Recurrent Neural Networks Are Universal Approximators2006
- Establishing Safety Criteria for Artificial Neural NetworksLecture Notes in Computer Science, 2003