Abstract
Safety is a critical concern for the next generation of autonomy that is likely to rely heavily on deep neural networks for perception and control. This paper proposes a method to repair unsafe ReLU DNNs in safety-critical systems using reachability analysis. Our repair method uses reachability analysis to calculate the unsafe reachable domain of a DNN, and then uses a novel loss function to construct its distance to the safe domain during the retraining process. Since subtle changes of the DNN parameters can cause unexpected performance degradation, we also present a minimal repair approach where the DNN deviation is minimized. Furthermore, we explore applications of our method to repair DNN agents in deep reinforcement learning (DRL) with seamless integration with learning algorithms. Our method is evaluated on the ACAS Xu benchmark and a rocket lander system against the state-of-the-art method ART. Experimental results show that our repair approach can generate provably safe DNNs on multiple safety specifications with negligible performance degradation, even in the absence of training data (Code is available online at https://github.com/Shaddadi/veritex.git).
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Brockman, G., et al.: Openai gym. arXiv preprint arXiv:1606.01540 (2016)
Goldberger, B., Katz, G., Adi, Y., Keshet, J.: Minimal modifications of deep neural networks using verification. In: LPAR, vol. 2020, p. 23 (2020)
Goodfellow, I.J., Shlens, J., Szegedy, C.: Explaining and harnessing adversarial examples. arXiv preprint arXiv:1412.6572 (2014)
Julian, K.D., Lopez, J., Brush, J.S., Owen, M.P., Kochenderfer, M.J.: Policy compression for aircraft collision avoidance systems. In: 2016 IEEE/AIAA 35th Digital Avionics Systems Conference (DASC), pp. 1–10. IEEE (2016)
Katz, G., Barrett, C., Dill, D.L., Julian, K., Kochenderfer, M.J.: Reluplex: an efficient SMT solver for verifying deep neural networks. In: Majumdar, R., Kunčak, V. (eds.) Reluplex: an efficient smt solver for verifying deep neural networks. LNCS, vol. 10426, pp. 97–117. Springer, Cham (2017). https://doi.org/10.1007/978-3-319-63387-9_5
Katz, G., et al.: The marabou framework for verification and analysis of deep neural networks. In: Dillig, I., Tasiran, S. (eds.) CAV 2019. LNCS, vol. 11561, pp. 443–452. Springer, Cham (2019). https://doi.org/10.1007/978-3-030-25540-4_26
Lillicrap, T.P., et al.: Continuous control with deep reinforcement learning. arXiv preprint arXiv:1509.02971 (2015)
Lin, X., Zhu, H., Samanta, R., Jagannathan, S.: Art: Abstraction refinement-guided training for provably correct neural networks. In: FMCAD, pp. 148–157 (2020)
Liu, C., Arnon, T., Lazarus, C., Strong, C., Barrett, C., Kochenderfer, M.J.: Algorithms for verifying deep neural networks. arXiv preprint arXiv:1903.06758 (2019)
Madry, A., Makelov, A., Schmidt, L., Tsipras, D., Vladu, A.: Towards deep learning models resistant to adversarial attacks. arXiv preprint arXiv:1706.06083 (2017)
Majd, K., Zhou, S., Amor, H.B., Fainekos, G., Sankaranarayanan, S.: Local repair of neural networks using optimization. arXiv preprint arXiv:2109.14041 (2021)
Mirman, M., Gehr, T., Vechev, M.: Differentiable abstract interpretation for provably robust neural networks. In: International Conference on Machine Learning, pp. 3578–3586 (2018)
Sohn, J., Kang, S., Yoo, S.: Search based repair of deep neural networks. arXiv preprint arXiv:1912.12463 (2019)
Tran, H.-D., et al.: NNV: the neural network verification tool for deep neural networks and learning-enabled cyber-physical systems. In: Lahiri, S.K., Wang, C. (eds.) CAV 2020. LNCS, vol. 12224, pp. 3–17. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-53288-8_1
Usman, M., Gopinath, D., Sun, Y., Noller, Y., Pasareanu, C.: NNrepair: constraint-based repair of neural network classifiers. arXiv preprint arXiv:2103.12535 (2021)
Wang, S., Pei, K., Whitehouse, J., Yang, J., Jana, S.: Formal security analysis of neural networks using symbolic intervals. In: 27th \(\{\)USENIX\(\}\) Security Symposium (\(\{\)USENIX\(\}\) Security 2018), pp. 1599–1614 (2018)
Wong, E., Kolter, Z.: Provable defenses against adversarial examples via the convex outer adversarial polytope. In: International Conference on Machine Learning, pp. 5286–5295. PMLR (2018)
Yang, X., Johnson, T.T., Tran, H.D., Yamaguchi, T., Hoxha, B., Prokhorov, D.: Reachability analysis of deep RELU neural networks using facet-vertex incidence. In: Proceedings of the 24th International Conference on Hybrid Systems: Computation and Control. HSCC 2021, Association for Computing Machinery, New York, NY, USA (2021). https://doi.org/10.1145/3447928.3456650
Yang, X., Yamaguchi, T., Tran, H.D., Hoxha, B., Johnson, T.T., Prokhorov, D.: Reachability analysis of convolutional neural networks. arXiv preprint arXiv:2106.12074 (2021)
Acknowledgements
The material presented in this paper is based upon work supported by the National Science Foundation (NSF) through grant numbers 1910017 and 2028001, the Defense Advanced Research Projects Agency (DARPA) under contract number FA8750-18-C-0089, and the Air Force Office of Scientific Research (AFOSR) under contract number FA9550-22-1-0019. Any opinions, findings, and conclusions or recommendations expressed in this paper are those of the authors and do not necessarily reflect the views of AFOSR, DARPA, or NSF.
Author information
Authors and Affiliations
Corresponding authors
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 Springer Nature Switzerland AG
About this paper
Cite this paper
Yang, X., Yamaguchi, T., Tran, HD., Hoxha, B., Johnson, T.T., Prokhorov, D. (2022). Neural Network Repair with Reachability Analysis. In: Bogomolov, S., Parker, D. (eds) Formal Modeling and Analysis of Timed Systems. FORMATS 2022. Lecture Notes in Computer Science, vol 13465. Springer, Cham. https://doi.org/10.1007/978-3-031-15839-1_13
Download citation
DOI: https://doi.org/10.1007/978-3-031-15839-1_13
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-15838-4
Online ISBN: 978-3-031-15839-1
eBook Packages: Computer ScienceComputer Science (R0)