Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox
Picture for Huan Zhang

Huan Zhang

Deep Image Destruction: A Comprehensive Study on Vulnerability of Deep Image-to-Image Models against Adversarial Attacks


Apr 30, 2021
Jun-Ho Choi, Huan Zhang, Jun-Hyuk Kim, Cho-Jui Hsieh, Jong-Seok Lee

Figure 1 for Deep Image Destruction: A Comprehensive Study on Vulnerability of Deep Image-to-Image Models against Adversarial Attacks
Figure 2 for Deep Image Destruction: A Comprehensive Study on Vulnerability of Deep Image-to-Image Models against Adversarial Attacks
Figure 3 for Deep Image Destruction: A Comprehensive Study on Vulnerability of Deep Image-to-Image Models against Adversarial Attacks
Figure 4 for Deep Image Destruction: A Comprehensive Study on Vulnerability of Deep Image-to-Image Models against Adversarial Attacks

  Access Paper or Ask Questions

Double Perturbation: On the Robustness of Robustness and Counterfactual Bias Evaluation


Apr 12, 2021
Chong Zhang, Jieyu Zhao, Huan Zhang, Kai-Wei Chang, Cho-Jui Hsieh

Figure 1 for Double Perturbation: On the Robustness of Robustness and Counterfactual Bias Evaluation
Figure 2 for Double Perturbation: On the Robustness of Robustness and Counterfactual Bias Evaluation
Figure 3 for Double Perturbation: On the Robustness of Robustness and Counterfactual Bias Evaluation
Figure 4 for Double Perturbation: On the Robustness of Robustness and Counterfactual Bias Evaluation
* NAACL 2021 
  Access Paper or Ask Questions

Fast Certified Robust Training via Better Initialization and Shorter Warmup


Apr 01, 2021
Zhouxing Shi, Yihan Wang, Huan Zhang, Jinfeng Yi, Cho-Jui Hsieh

Figure 1 for Fast Certified Robust Training via Better Initialization and Shorter Warmup
Figure 2 for Fast Certified Robust Training via Better Initialization and Shorter Warmup
Figure 3 for Fast Certified Robust Training via Better Initialization and Shorter Warmup
Figure 4 for Fast Certified Robust Training via Better Initialization and Shorter Warmup

  Access Paper or Ask Questions

Beta-CROWN: Efficient Bound Propagation with Per-neuron Split Constraints for Complete and Incomplete Neural Network Verification


Mar 11, 2021
Shiqi Wang, Huan Zhang, Kaidi Xu, Xue Lin, Suman Jana, Cho-Jui Hsieh, J. Zico Kolter

Figure 1 for Beta-CROWN: Efficient Bound Propagation with Per-neuron Split Constraints for Complete and Incomplete Neural Network Verification
Figure 2 for Beta-CROWN: Efficient Bound Propagation with Per-neuron Split Constraints for Complete and Incomplete Neural Network Verification
Figure 3 for Beta-CROWN: Efficient Bound Propagation with Per-neuron Split Constraints for Complete and Incomplete Neural Network Verification
Figure 4 for Beta-CROWN: Efficient Bound Propagation with Per-neuron Split Constraints for Complete and Incomplete Neural Network Verification
* Shiqi Wang, Huan Zhang and Kaidi Xu contributed equally 
  Access Paper or Ask Questions

Does deep machine vision have just noticeable difference (JND)?


Feb 16, 2021
Jian Jin, Xingxing Zhang, Xin Fu, Huan Zhang, Weisi Lin, Jian Lou, Yao Zhao

Figure 1 for Does deep machine vision have just noticeable difference (JND)?
Figure 2 for Does deep machine vision have just noticeable difference (JND)?
Figure 3 for Does deep machine vision have just noticeable difference (JND)?
Figure 4 for Does deep machine vision have just noticeable difference (JND)?
* 9 pages, 8 figures 
  Access Paper or Ask Questions

Robust Reinforcement Learning on State Observations with Learned Optimal Adversary


Jan 21, 2021
Huan Zhang, Hongge Chen, Duane Boning, Cho-Jui Hsieh

Figure 1 for Robust Reinforcement Learning on State Observations with Learned Optimal Adversary
Figure 2 for Robust Reinforcement Learning on State Observations with Learned Optimal Adversary
Figure 3 for Robust Reinforcement Learning on State Observations with Learned Optimal Adversary
Figure 4 for Robust Reinforcement Learning on State Observations with Learned Optimal Adversary
* Accepted by ICLR 2021. Huan Zhang and Hongge Chen contributed equally 
  Access Paper or Ask Questions

Fast and Complete: Enabling Complete Neural Network Verification with Rapid and Massively Parallel Incomplete Verifiers


Nov 27, 2020
Kaidi Xu, Huan Zhang, Shiqi Wang, Yihan Wang, Suman Jana, Xue Lin, Cho-Jui Hsieh

Figure 1 for Fast and Complete: Enabling Complete Neural Network Verification with Rapid and Massively Parallel Incomplete Verifiers
Figure 2 for Fast and Complete: Enabling Complete Neural Network Verification with Rapid and Massively Parallel Incomplete Verifiers
Figure 3 for Fast and Complete: Enabling Complete Neural Network Verification with Rapid and Massively Parallel Incomplete Verifiers
Figure 4 for Fast and Complete: Enabling Complete Neural Network Verification with Rapid and Massively Parallel Incomplete Verifiers

  Access Paper or Ask Questions

An Efficient Adversarial Attack for Tree Ensembles


Oct 22, 2020
Chong Zhang, Huan Zhang, Cho-Jui Hsieh

Figure 1 for An Efficient Adversarial Attack for Tree Ensembles
Figure 2 for An Efficient Adversarial Attack for Tree Ensembles
Figure 3 for An Efficient Adversarial Attack for Tree Ensembles
Figure 4 for An Efficient Adversarial Attack for Tree Ensembles
* NeurIPS 2020 
  Access Paper or Ask Questions

On $\ell_p$-norm Robustness of Ensemble Stumps and Trees


Sep 29, 2020
Yihan Wang, Huan Zhang, Hongge Chen, Duane Boning, Cho-Jui Hsieh

Figure 1 for On $\ell_p$-norm Robustness of Ensemble Stumps and Trees
Figure 2 for On $\ell_p$-norm Robustness of Ensemble Stumps and Trees
Figure 3 for On $\ell_p$-norm Robustness of Ensemble Stumps and Trees
Figure 4 for On $\ell_p$-norm Robustness of Ensemble Stumps and Trees
* ICML 2020 
  Access Paper or Ask Questions

The Limit of the Batch Size


Jun 15, 2020
Yang You, Yuhui Wang, Huan Zhang, Zhao Zhang, James Demmel, Cho-Jui Hsieh

Figure 1 for The Limit of the Batch Size
Figure 2 for The Limit of the Batch Size
Figure 3 for The Limit of the Batch Size
Figure 4 for The Limit of the Batch Size

  Access Paper or Ask Questions

Spanning Attack: Reinforce Black-box Attacks with Unlabeled Data


May 11, 2020
Lu Wang, Huan Zhang, Jinfeng Yi, Cho-Jui Hsieh, Yuan Jiang

Figure 1 for Spanning Attack: Reinforce Black-box Attacks with Unlabeled Data
Figure 2 for Spanning Attack: Reinforce Black-box Attacks with Unlabeled Data
Figure 3 for Spanning Attack: Reinforce Black-box Attacks with Unlabeled Data
Figure 4 for Spanning Attack: Reinforce Black-box Attacks with Unlabeled Data

  Access Paper or Ask Questions

Towards Non-task-specific Distillation of BERT via Sentence Representation Approximation


Apr 07, 2020
Bowen Wu, Huan Zhang, Mengyuan Li, Zongsheng Wang, Qihang Feng, Junhong Huang, Baoxun Wang

Figure 1 for Towards Non-task-specific Distillation of BERT via Sentence Representation Approximation
Figure 2 for Towards Non-task-specific Distillation of BERT via Sentence Representation Approximation
Figure 3 for Towards Non-task-specific Distillation of BERT via Sentence Representation Approximation
Figure 4 for Towards Non-task-specific Distillation of BERT via Sentence Representation Approximation

  Access Paper or Ask Questions

Robust Deep Reinforcement Learning against Adversarial Perturbations on Observations


Mar 19, 2020
Huan Zhang, Hongge Chen, Chaowei Xiao, Bo Li, Duane Boning, Cho-Jui Hsieh

Figure 1 for Robust Deep Reinforcement Learning against Adversarial Perturbations on Observations
Figure 2 for Robust Deep Reinforcement Learning against Adversarial Perturbations on Observations
Figure 3 for Robust Deep Reinforcement Learning against Adversarial Perturbations on Observations
Figure 4 for Robust Deep Reinforcement Learning against Adversarial Perturbations on Observations
* Huan Zhang and Hongge Chen contributed equally 
  Access Paper or Ask Questions

Automatic Perturbation Analysis on General Computational Graphs


Feb 28, 2020
Kaidi Xu, Zhouxing Shi, Huan Zhang, Minlie Huang, Kai-Wei Chang, Bhavya Kailkhura, Xue Lin, Cho-Jui Hsieh

Figure 1 for Automatic Perturbation Analysis on General Computational Graphs
Figure 2 for Automatic Perturbation Analysis on General Computational Graphs
Figure 3 for Automatic Perturbation Analysis on General Computational Graphs
Figure 4 for Automatic Perturbation Analysis on General Computational Graphs

  Access Paper or Ask Questions

Robustness Verification for Transformers


Feb 16, 2020
Zhouxing Shi, Huan Zhang, Kai-Wei Chang, Minlie Huang, Cho-Jui Hsieh

Figure 1 for Robustness Verification for Transformers
Figure 2 for Robustness Verification for Transformers
Figure 3 for Robustness Verification for Transformers
Figure 4 for Robustness Verification for Transformers
* ICLR 2020 
  Access Paper or Ask Questions

MACER: Attack-free and Scalable Robust Training via Maximizing Certified Radius


Feb 15, 2020
Runtian Zhai, Chen Dan, Di He, Huan Zhang, Boqing Gong, Pradeep Ravikumar, Cho-Jui Hsieh, Liwei Wang

Figure 1 for MACER: Attack-free and Scalable Robust Training via Maximizing Certified Radius
Figure 2 for MACER: Attack-free and Scalable Robust Training via Maximizing Certified Radius
Figure 3 for MACER: Attack-free and Scalable Robust Training via Maximizing Certified Radius
Figure 4 for MACER: Attack-free and Scalable Robust Training via Maximizing Certified Radius
* In ICLR 2020. 20 Pages 
  Access Paper or Ask Questions

Robust Triple-Matrix-Recovery-Based Auto-Weighted Label Propagation for Classification


Nov 20, 2019
Huan Zhang, Zhao Zhang, Mingbo Zhao, Qiaolin Ye, Min Zhang, Meng Wang

Figure 1 for Robust Triple-Matrix-Recovery-Based Auto-Weighted Label Propagation for Classification
Figure 2 for Robust Triple-Matrix-Recovery-Based Auto-Weighted Label Propagation for Classification
Figure 3 for Robust Triple-Matrix-Recovery-Based Auto-Weighted Label Propagation for Classification
Figure 4 for Robust Triple-Matrix-Recovery-Based Auto-Weighted Label Propagation for Classification
* Accepted by IEEE TNNNLS 
  Access Paper or Ask Questions

Reducing Sentiment Bias in Language Models via Counterfactual Evaluation


Nov 08, 2019
Po-Sen Huang, Huan Zhang, Ray Jiang, Robert Stanforth, Johannes Welbl, Jack Rae, Vishal Maini, Dani Yogatama, Pushmeet Kohli

Figure 1 for Reducing Sentiment Bias in Language Models via Counterfactual Evaluation
Figure 2 for Reducing Sentiment Bias in Language Models via Counterfactual Evaluation
Figure 3 for Reducing Sentiment Bias in Language Models via Counterfactual Evaluation
Figure 4 for Reducing Sentiment Bias in Language Models via Counterfactual Evaluation

  Access Paper or Ask Questions

Enhancing Certifiable Robustness via a Deep Model Ensemble


Oct 31, 2019
Huan Zhang, Minhao Cheng, Cho-Jui Hsieh

Figure 1 for Enhancing Certifiable Robustness via a Deep Model Ensemble
Figure 2 for Enhancing Certifiable Robustness via a Deep Model Ensemble
Figure 3 for Enhancing Certifiable Robustness via a Deep Model Ensemble
Figure 4 for Enhancing Certifiable Robustness via a Deep Model Ensemble
* This is an extended version of ICLR 2019 Safe Machine Learning Workshop (SafeML) paper, "RobBoost: A provable approach to boost the robustness of deep model ensemble". May 6, 2019, New Orleans, LA, USA 
  Access Paper or Ask Questions

MemeFaceGenerator: Adversarial Synthesis of Chinese Meme-face from Natural Sentences


Aug 14, 2019
Yifu Chen, Zongsheng Wang, Bowen Wu, Mengyuan Li, Huan Zhang, Lin Ma, Feng Liu, Qihang Feng, Baoxun Wang

Figure 1 for MemeFaceGenerator: Adversarial Synthesis of Chinese Meme-face from Natural Sentences
Figure 2 for MemeFaceGenerator: Adversarial Synthesis of Chinese Meme-face from Natural Sentences
Figure 3 for MemeFaceGenerator: Adversarial Synthesis of Chinese Meme-face from Natural Sentences
Figure 4 for MemeFaceGenerator: Adversarial Synthesis of Chinese Meme-face from Natural Sentences

  Access Paper or Ask Questions

Defending Against Adversarial Attacks Using Random Forests


Jun 16, 2019
Yifan Ding, Liqiang Wang, Huan Zhang, Jinfeng Yi, Deliang Fan, Boqing Gong

Figure 1 for Defending Against Adversarial Attacks Using Random Forests
Figure 2 for Defending Against Adversarial Attacks Using Random Forests
Figure 3 for Defending Against Adversarial Attacks Using Random Forests
Figure 4 for Defending Against Adversarial Attacks Using Random Forests

  Access Paper or Ask Questions

Robustness Verification of Tree-based Models


Jun 15, 2019
Hongge Chen, Huan Zhang, Si Si, Yang Li, Duane Boning, Cho-Jui Hsieh

Figure 1 for Robustness Verification of Tree-based Models
Figure 2 for Robustness Verification of Tree-based Models
Figure 3 for Robustness Verification of Tree-based Models
Figure 4 for Robustness Verification of Tree-based Models
* Hongge Chen and Huan Zhang contributed equally 
  Access Paper or Ask Questions

Towards Stable and Efficient Training of Verifiably Robust Neural Networks


Jun 14, 2019
Huan Zhang, Hongge Chen, Chaowei Xiao, Bo Li, Duane Boning, Cho-Jui Hsieh

Figure 1 for Towards Stable and Efficient Training of Verifiably Robust Neural Networks
Figure 2 for Towards Stable and Efficient Training of Verifiably Robust Neural Networks
Figure 3 for Towards Stable and Efficient Training of Verifiably Robust Neural Networks
Figure 4 for Towards Stable and Efficient Training of Verifiably Robust Neural Networks

  Access Paper or Ask Questions

Provably Robust Deep Learning via Adversarially Trained Smoothed Classifiers


Jun 12, 2019
Hadi Salman, Greg Yang, Jerry Li, Pengchuan Zhang, Huan Zhang, Ilya Razenshteyn, Sebastien Bubeck

Figure 1 for Provably Robust Deep Learning via Adversarially Trained Smoothed Classifiers
Figure 2 for Provably Robust Deep Learning via Adversarially Trained Smoothed Classifiers
Figure 3 for Provably Robust Deep Learning via Adversarially Trained Smoothed Classifiers
Figure 4 for Provably Robust Deep Learning via Adversarially Trained Smoothed Classifiers
* 27 pages 
  Access Paper or Ask Questions

Evaluating Robustness of Deep Image Super-Resolution against Adversarial Attacks


Apr 12, 2019
Jun-Ho Choi, Huan Zhang, Jun-Hyuk Kim, Cho-Jui Hsieh, Jong-Seok Lee

Figure 1 for Evaluating Robustness of Deep Image Super-Resolution against Adversarial Attacks
Figure 2 for Evaluating Robustness of Deep Image Super-Resolution against Adversarial Attacks
Figure 3 for Evaluating Robustness of Deep Image Super-Resolution against Adversarial Attacks
Figure 4 for Evaluating Robustness of Deep Image Super-Resolution against Adversarial Attacks

  Access Paper or Ask Questions

Second Rethinking of Network Pruning in the Adversarial Setting


Mar 29, 2019
Shaokai Ye, Kaidi Xu, Sijia Liu, Hao Cheng, Jan-Henrik Lambrechts, Huan Zhang, Aojun Zhou, Kaisheng Ma, Yanzhi Wang, Xue Lin

Figure 1 for Second Rethinking of Network Pruning in the Adversarial Setting
Figure 2 for Second Rethinking of Network Pruning in the Adversarial Setting
Figure 3 for Second Rethinking of Network Pruning in the Adversarial Setting
Figure 4 for Second Rethinking of Network Pruning in the Adversarial Setting

  Access Paper or Ask Questions