Picture for Xilie Xu

Xilie Xu

Privacy-Preserving Low-Rank Adaptation for Latent Diffusion Models

Add code
Feb 19, 2024
Viaarxiv icon

AutoLoRa: A Parameter-Free Automated Robust Fine-Tuning Framework

Add code
Oct 03, 2023
Viaarxiv icon

Enhancing Adversarial Contrastive Learning via Adversarial Invariant Regularization

Add code
Apr 30, 2023
Viaarxiv icon

Efficient Adversarial Contrastive Learning via Robustness-Aware Coreset Selection

Add code
Feb 08, 2023
Viaarxiv icon

Adversarial Attacks and Defense for Non-Parametric Two-Sample Tests

Add code
Feb 07, 2022
Figure 1 for Adversarial Attacks and Defense for Non-Parametric Two-Sample Tests
Figure 2 for Adversarial Attacks and Defense for Non-Parametric Two-Sample Tests
Figure 3 for Adversarial Attacks and Defense for Non-Parametric Two-Sample Tests
Figure 4 for Adversarial Attacks and Defense for Non-Parametric Two-Sample Tests
Viaarxiv icon

NoiLIn: Do Noisy Labels Always Hurt Adversarial Training?

Add code
May 31, 2021
Figure 1 for NoiLIn: Do Noisy Labels Always Hurt Adversarial Training?
Figure 2 for NoiLIn: Do Noisy Labels Always Hurt Adversarial Training?
Figure 3 for NoiLIn: Do Noisy Labels Always Hurt Adversarial Training?
Figure 4 for NoiLIn: Do Noisy Labels Always Hurt Adversarial Training?
Viaarxiv icon

Guided Interpolation for Adversarial Training

Add code
Feb 15, 2021
Figure 1 for Guided Interpolation for Adversarial Training
Figure 2 for Guided Interpolation for Adversarial Training
Figure 3 for Guided Interpolation for Adversarial Training
Figure 4 for Guided Interpolation for Adversarial Training
Viaarxiv icon

Attacks Which Do Not Kill Training Make Adversarial Learning Stronger

Add code
Feb 26, 2020
Figure 1 for Attacks Which Do Not Kill Training Make Adversarial Learning Stronger
Figure 2 for Attacks Which Do Not Kill Training Make Adversarial Learning Stronger
Figure 3 for Attacks Which Do Not Kill Training Make Adversarial Learning Stronger
Figure 4 for Attacks Which Do Not Kill Training Make Adversarial Learning Stronger
Viaarxiv icon