Incremental Randomized Smoothing Certification

TL;DR

This paper introduces IRS, an incremental certification method for randomized smoothing that reuses previous guarantees to efficiently certify modified neural networks, significantly reducing computational costs.

Contribution

IRS is the first approach to reuse certification guarantees for modified smoothed models, enabling faster certification with minimal samples.

Findings

Up to 3x speedup in certification time

Maintains strong robustness guarantees

Effective for models after modifications like quantization or pruning

Abstract

Randomized smoothing-based certification is an effective approach for obtaining robustness certificates of deep neural networks (DNNs) against adversarial attacks. This method constructs a smoothed DNN model and certifies its robustness through statistical sampling, but it is computationally expensive, especially when certifying with a large number of samples. Furthermore, when the smoothed model is modified (e.g., quantized or pruned), certification guarantees may not hold for the modified DNN, and recertifying from scratch can be prohibitively expensive. We present the first approach for incremental robustness certification for randomized smoothing, IRS. We show how to reuse the certification guarantees for the original smoothed model to certify an approximated model with very few samples. IRS significantly reduces the computational cost of certifying modified DNNs while maintaining strong robustness guarantees. We experimentally demonstrate the effectiveness of our approach, showing up to 3x certification speedup over the certification that applies randomized smoothing of the approximate model from scratch.