Timezone: »

Automatic Perturbation Analysis for Scalable Certified Robustness and Beyond
Kaidi Xu · Zhouxing Shi · Huan Zhang · Yihan Wang · Kai-Wei Chang · Minlie Huang · Bhavya Kailkhura · Xue Lin · Cho-Jui Hsieh

Thu Dec 10 09:00 AM -- 11:00 AM (PST) @ Poster Session 5 #1593

Linear relaxation based perturbation analysis (LiRPA) for neural networks, which computes provable linear bounds of output neurons given a certain amount of input perturbation, has become a core component in robustness verification and certified defense. The majority of LiRPA-based methods focus on simple feed-forward networks and need particular manual derivations and implementations when extended to other architectures. In this paper, we develop an automatic framework to enable perturbation analysis on any neural network structures, by generalizing existing LiRPA algorithms such as CROWN to operate on general computational graphs. The flexibility, differentiability and ease of use of our framework allow us to obtain state-of-the-art results on LiRPA based certified defense for fairly complicated networks like DenseNet, ResNeXt and Transformer that are not supported by prior works. Our framework also enables loss fusion, a technique that significantly reduces the computational complexity of LiRPA for certified defense. For the first time, we demonstrate LiRPA based certified defense on Tiny ImageNet and Downscaled ImageNet where previous approaches cannot scale to due to the relatively large number of classes. Our work also yields an open-source library for the community to apply LiRPA to areas beyond adversarial robustness without much LiRPA expertise, e.g., we create a neural network with a provably flat optimization landscape by applying LiRPA to network parameters and considering perturbations on model weights. Our open source library is available at https://github.com/KaidiXu/auto_LiRPA.

Author Information

Kaidi Xu (Northeastern University)
Zhouxing Shi (UCLA)
Huan Zhang (UCLA)
Yihan Wang (UCLA)
Kai-Wei Chang (UCLA)
Minlie Huang (Tsinghua University)
Bhavya Kailkhura (Lawrence Livermore National Laboratory)
Xue Lin (Northeastern University)
Cho-Jui Hsieh (UCLA)

More from the Same Authors