Picture for Yuefeng Chen

Yuefeng Chen

S-Eval: Automatic and Adaptive Test Generation for Benchmarking Safety Evaluation of Large Language Models

May 28, 2024
Viaarxiv icon

Enhancing Few-shot CLIP with Semantic-Aware Fine-Tuning

Add code
Nov 09, 2023
Viaarxiv icon

Revisiting and Exploring Efficient Fast Adversarial Training via LAW: Lipschitz Regularization and Auto Weight Averaging

Aug 22, 2023
Figure 1 for Revisiting and Exploring Efficient Fast Adversarial Training via LAW: Lipschitz Regularization and Auto Weight Averaging
Figure 2 for Revisiting and Exploring Efficient Fast Adversarial Training via LAW: Lipschitz Regularization and Auto Weight Averaging
Figure 3 for Revisiting and Exploring Efficient Fast Adversarial Training via LAW: Lipschitz Regularization and Auto Weight Averaging
Figure 4 for Revisiting and Exploring Efficient Fast Adversarial Training via LAW: Lipschitz Regularization and Auto Weight Averaging
Viaarxiv icon

COCO-O: A Benchmark for Object Detectors under Natural Distribution Shifts

Add code
Aug 02, 2023
Viaarxiv icon

Robust Automatic Speech Recognition via WavAugment Guided Phoneme Adversarial Training

Add code
Jul 24, 2023
Figure 1 for Robust Automatic Speech Recognition via WavAugment Guided Phoneme Adversarial Training
Figure 2 for Robust Automatic Speech Recognition via WavAugment Guided Phoneme Adversarial Training
Figure 3 for Robust Automatic Speech Recognition via WavAugment Guided Phoneme Adversarial Training
Viaarxiv icon

ImageNet-E: Benchmarking Neural Network Robustness via Attribute Editing

Add code
Mar 30, 2023
Figure 1 for ImageNet-E: Benchmarking Neural Network Robustness via Attribute Editing
Figure 2 for ImageNet-E: Benchmarking Neural Network Robustness via Attribute Editing
Figure 3 for ImageNet-E: Benchmarking Neural Network Robustness via Attribute Editing
Figure 4 for ImageNet-E: Benchmarking Neural Network Robustness via Attribute Editing
Viaarxiv icon

TransAudio: Towards the Transferable Adversarial Audio Attack via Learning Contextualized Perturbations

Add code
Mar 28, 2023
Figure 1 for TransAudio: Towards the Transferable Adversarial Audio Attack via Learning Contextualized Perturbations
Figure 2 for TransAudio: Towards the Transferable Adversarial Audio Attack via Learning Contextualized Perturbations
Figure 3 for TransAudio: Towards the Transferable Adversarial Audio Attack via Learning Contextualized Perturbations
Figure 4 for TransAudio: Towards the Transferable Adversarial Audio Attack via Learning Contextualized Perturbations
Viaarxiv icon

Information-containing Adversarial Perturbation for Combating Facial Manipulation Systems

Mar 21, 2023
Figure 1 for Information-containing Adversarial Perturbation for Combating Facial Manipulation Systems
Figure 2 for Information-containing Adversarial Perturbation for Combating Facial Manipulation Systems
Figure 3 for Information-containing Adversarial Perturbation for Combating Facial Manipulation Systems
Figure 4 for Information-containing Adversarial Perturbation for Combating Facial Manipulation Systems
Viaarxiv icon

A Comprehensive Study on Robustness of Image Classification Models: Benchmarking and Rethinking

Add code
Feb 28, 2023
Figure 1 for A Comprehensive Study on Robustness of Image Classification Models: Benchmarking and Rethinking
Figure 2 for A Comprehensive Study on Robustness of Image Classification Models: Benchmarking and Rethinking
Figure 3 for A Comprehensive Study on Robustness of Image Classification Models: Benchmarking and Rethinking
Figure 4 for A Comprehensive Study on Robustness of Image Classification Models: Benchmarking and Rethinking
Viaarxiv icon

Rethinking Out-of-Distribution Detection From a Human-Centric Perspective

Nov 30, 2022
Figure 1 for Rethinking Out-of-Distribution Detection From a Human-Centric Perspective
Figure 2 for Rethinking Out-of-Distribution Detection From a Human-Centric Perspective
Figure 3 for Rethinking Out-of-Distribution Detection From a Human-Centric Perspective
Figure 4 for Rethinking Out-of-Distribution Detection From a Human-Centric Perspective
Viaarxiv icon