93 citations · 629 across the 30 of their papers we have counts for
30 papers
Can Agents Run Relay Race with Strangers? Generalization of RL to Out-of-Distribution Trajectories
Li-Cheng Lan, Huan Zhang, Cho-Jui Hsieh
In this paper, we define, evaluate, and improve the ``relay-generalization'' performance of reinforcement learning (RL) agents on the out-of-distribution ``controllable'' states. I…
Are AlphaZero-like Agents Robust to Adversarial Perturbations?
Li-Cheng Lan, Huan Zhang, Ti-Rong Wu +3
The success of AlphaZero (AZ) has demonstrated that neural-network-based Go AIs can surpass human performance by a large margin. Given that the state space of Go is extremely large…
Efficiently Computing Local Lipschitz Constants of Neural Networks via Bound Propagation
Zhouxing Shi, Yihan Wang, Huan Zhang +2
Lipschitz constants are connected to many properties of neural networks, such as robustness, fairness, and generalization. Existing methods for computing Lipschitz constants either…
General Cutting Planes for Bound-Propagation-Based Neural Network Verification
Huan Zhang, Shiqi Wang, Kaidi Xu +5
Bound propagation methods, when combined with branch and bound, are among the most effective methods to formally verify properties of deep neural networks such as correctness, robu…
On the Robustness of Safe Reinforcement Learning under Observational Perturbations
Zuxin Liu, Zijian Guo, Zhepeng Cen +4
Safe reinforcement learning (RL) trains a policy to maximize the task reward while satisfying safety constraints. While prior works focus on the performance optimality, we find tha…
COPA: Certifying Robust Policies for Offline Reinforcement Learning against Poisoning Attacks
Fan Wu, Linyi Li, Chejian Xu +5
As reinforcement learning (RL) has achieved near human-level performance in a variety of tasks, its robustness has raised great attention. While a vast body of research has explore…
Temporal Shuffling for Defending Deep Action Recognition Models against Adversarial Attacks
Jaehui Hwang, Huan Zhang, Jun-Ho Choi +2
Recently, video-based action recognition methods using convolutional neural networks (CNNs) achieve remarkable recognition performance. However, there is still lack of understandin…
Training Certifiably Robust Neural Networks with Efficient Local Lipschitz Bounds
Yujia Huang, Huan Zhang, Yuanyuan Shi +2
Certified robustness is a desirable property for deep neural networks in safety-critical applications, and popular training algorithms can certify robustness of a neural network by…
Improving Robustness of Reinforcement Learning for Power System Control with Adversarial Training
Alexander Pan, Yongkyun Lee, Huan Zhang +2
Due to the proliferation of renewable energy and its intrinsic intermittency and stochasticity, current power systems face severe operational challenges. Data-driven decision-makin…
Double Perturbation: On the Robustness of Robustness and Counterfactual Bias Evaluation
Chong Zhang, Jieyu Zhao, Huan Zhang +2
Robustness and counterfactual bias are usually evaluated on a test dataset. However, are these evaluations robust? If the test dataset is perturbed slightly, will the evaluation re…
Fast Certified Robust Training with Short Warmup
Zhouxing Shi, Yihan Wang, Huan Zhang +2
Recently, bound propagation based certified robust training methods have been proposed for training neural networks with certifiable robustness guarantees. Despite that state-of-th…
Beta-CROWN: Efficient Bound Propagation with Per-neuron Split Constraints for Complete and Incomplete Neural Network Robustness Verification
Shiqi Wang, Huan Zhang, Kaidi Xu +4
Bound propagation based incomplete neural network verifiers such as CROWN are very efficient and can significantly accelerate branch-and-bound (BaB) based complete verification of…
Just Noticeable Difference for Deep Machine Vision
Jian Jin, Xingxing Zhang, Xin Fu +4
As an important perceptual characteristic of the Human Visual System (HVS), the Just Noticeable Difference (JND) has been studied for decades with image and video processing (e.g.,…
Robust Reinforcement Learning on State Observations with Learned Optimal Adversary
Huan Zhang, Hongge Chen, Duane Boning +1
We study the robustness of reinforcement learning (RL) with adversarially perturbed state observations, which aligns with the setting of many adversarial attacks to deep reinforcem…
Fast and Complete: Enabling Complete Neural Network Verification with Rapid and Massively Parallel Incomplete Verifiers
Kaidi Xu, Huan Zhang, Shiqi Wang +4
Formal verification of neural networks (NNs) is a challenging and important problem. Existing efficient complete solvers typically require the branch-and-bound (BaB) process, which…
An Efficient Adversarial Attack for Tree Ensembles
Chong Zhang, Huan Zhang, Cho-Jui Hsieh
We study the problem of efficient adversarial attacks on tree based ensembles such as gradient boosting decision trees (GBDTs) and random forests (RFs). Since these models are non-…
On -norm Robustness of Ensemble Stumps and Trees
Yihan Wang, Huan Zhang, Hongge Chen +2
Recent papers have demonstrated that ensemble stumps and trees could be vulnerable to small input perturbations, so robustness verification and defense for those models have become…
The Limit of the Batch Size
Yang You, Yuhui Wang, Huan Zhang +3
Large-batch training is an efficient approach for current distributed deep learning systems. It has enabled researchers to reduce the ImageNet/ResNet-50 training from 29 hours to a…
Spanning Attack: Reinforce Black-box Attacks with Unlabeled Data
Lu Wang, Huan Zhang, Jinfeng Yi +2
Adversarial black-box attacks aim to craft adversarial perturbations by querying input-output pairs of machine learning models. They are widely used to evaluate the robustness of p…
Robust Deep Reinforcement Learning against Adversarial Perturbations on State Observations
Huan Zhang, Hongge Chen, Chaowei Xiao +4
A deep reinforcement learning (DRL) agent observes its states through observations, which may contain natural measurement errors or adversarial noises. Since the observations devia…
Automatic Perturbation Analysis for Scalable Certified Robustness and Beyond
Kaidi Xu, Zhouxing Shi, Huan Zhang +6
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…
Advbox: a toolbox to generate adversarial examples that fool neural networks
Dou Goodman, Hao Xin, Wang Yang +3
In recent years, neural networks have been extensively deployed for computer vision tasks, particularly visual classification problems, where new algorithms reported to achieve or…
MACER: Attack-free and Scalable Robust Training via Maximizing Certified Radius
Runtian Zhai, Chen Dan, Di He +5
Adversarial training is one of the most popular ways to learn robust models but is usually attack-dependent and time costly. In this paper, we propose the MACER algorithm, which le…
Enhancing Certifiable Robustness via a Deep Model Ensemble
Huan Zhang, Minhao Cheng, Cho-Jui Hsieh
We propose an algorithm to enhance certified robustness of a deep model ensemble by optimally weighting each base model. Unlike previous works on using ensembles to empirically imp…