Shrink & Cert: Bi-level Optimization for Certified Robustness - CentraleSupélec Access content directly
Conference Papers Year : 2023

Shrink & Cert: Bi-level Optimization for Certified Robustness

Kavya Gupta
  • Function : Author
  • PersonId : 1089038
Sagar Verma
  • Function : Author
  • PersonId : 743250
  • IdHAL : versag


In this paper, we advance the concept of shrinking weights to train certifiably robust models from the fresh perspective of gradient-based bi-level optimization. Lack of robustness against adversarial attacks remains a challenge in safety-critical applications. Many attempts have been made in literature which only provide empirical verification of the defenses to certain attacks and can be easily broken. Methods in other lines of work can only develop certified guarantees of the model robustness in limited scenarios and are computationally expensive. We present a weight shrinkage formulation that is computationally inexpensive and can be solved as a simple first-order optimization problem. We show that model trained with our method has lower Lipschitz bounds in each layer, which directly provides formal guarantees on the certified robustness. We demonstrate that our approach, Shrink & Cert (SaC) achieves provably robust networks which simultaneously give excellent standard and robust accuracy. We demonstrate the success of our approach on CIFAR-10 and ImageNet datasets and compare them with existing robust training techniques. Code : https: //
Fichier principal
Vignette du fichier
advml_frontiers2023_Shrink_cert.pdf (276.45 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-04163747 , version 1 (17-07-2023)


  • HAL Id : hal-04163747 , version 1


Kavya Gupta, Sagar Verma. Shrink & Cert: Bi-level Optimization for Certified Robustness. New Frontiers in Adversarial Machine Learning, ICML, PMLR, Jul 2023, Honululu, United States. ⟨hal-04163747⟩
63 View
72 Download


Gmail Facebook X LinkedIn More