Abstract | ||
---|---|---|
Verifying the robustness property of a general Rectified Linear Unit (ReLU) network is an NP-complete problem [Katz, Barrett, Dill, Julian and Kochenderfer CAV17]. Although finding the exact minimum adversarial distortion is hard, giving a certified lower bound of the minimum distortion is possible. Current available methods of computing such a bound are either time-consuming or delivering low quality bounds that are too loose to be useful. this paper, we exploit the special structure of ReLU networks and provide two computationally efficient algorithms (Fast-Lin and Fast-Lip) that are able to certify non-trivial lower bounds of minimum distortions, by bounding the ReLU units with appropriate linear functions (Fast-Lin), or by bounding the local Lipschitz constant (Fast-Lip). Experiments show that (1) our proposed methods deliver bounds close to (the gap is 2-3X) exact minimum distortion found by Reluplex in small MNIST networks while our algorithms are more than 10,000 times faster; (2) our methods deliver similar quality of bounds (the gap is within 35% and usually around 10%; sometimes our bounds are even better) for larger networks compared to the methods based on solving linear programming problems but our algorithms are 33-14,000 times faster; (3) our method is capable of solving large MNIST and CIFAR networks up to 7 layers with more than 10,000 neurons within tens of seconds on a single CPU core. In addition, we show that, in fact, there is no polynomial time algorithm that can approximately find the minimum $ell_1$ adversarial distortion of a ReLU network with a $0.99ln n$ approximation ratio unless $mathsf{NP}$=$mathsf{P}$, where $n$ is the number of neurons in the network. |
Year | Venue | DocType |
---|---|---|
2018 | ICML | Conference |
Volume | Citations | PageRank |
abs/1804.09699 | 19 | 0.58 |
References | Authors | |
35 | 8 |
Name | Order | Citations | PageRank |
---|---|---|---|
Tsui-Wei Weng | 1 | 75 | 7.35 |
Huan Zhang | 2 | 327 | 23.01 |
Hongge Chen | 3 | 36 | 4.58 |
Zhao Song | 4 | 177 | 25.62 |
Cho-Jui Hsieh | 5 | 5034 | 291.05 |
Duane Boning | 6 | 201 | 49.37 |
Inderjit S. Dhillon | 7 | 7601 | 450.15 |
Luca Daniel | 8 | 497 | 50.96 |