Picture for Amin Ghiasi

Amin Ghiasi

What do Vision Transformers Learn? A Visual Exploration

Add code
Dec 13, 2022
Figure 1 for What do Vision Transformers Learn? A Visual Exploration
Figure 2 for What do Vision Transformers Learn? A Visual Exploration
Figure 3 for What do Vision Transformers Learn? A Visual Exploration
Figure 4 for What do Vision Transformers Learn? A Visual Exploration
Viaarxiv icon

Adaptive Weight Decay: On The Fly Weight Decay Tuning for Improving Robustness

Sep 30, 2022
Figure 1 for Adaptive Weight Decay: On The Fly Weight Decay Tuning for Improving Robustness
Figure 2 for Adaptive Weight Decay: On The Fly Weight Decay Tuning for Improving Robustness
Figure 3 for Adaptive Weight Decay: On The Fly Weight Decay Tuning for Improving Robustness
Figure 4 for Adaptive Weight Decay: On The Fly Weight Decay Tuning for Improving Robustness
Viaarxiv icon

Plug-In Inversion: Model-Agnostic Inversion for Vision with Data Augmentations

Add code
Jan 31, 2022
Figure 1 for Plug-In Inversion: Model-Agnostic Inversion for Vision with Data Augmentations
Figure 2 for Plug-In Inversion: Model-Agnostic Inversion for Vision with Data Augmentations
Figure 3 for Plug-In Inversion: Model-Agnostic Inversion for Vision with Data Augmentations
Figure 4 for Plug-In Inversion: Model-Agnostic Inversion for Vision with Data Augmentations
Viaarxiv icon

DP-InstaHide: Provably Defusing Poisoning and Backdoor Attacks with Differentially Private Data Augmentations

Mar 02, 2021
Viaarxiv icon

Strong Data Augmentation Sanitizes Poisoning and Backdoor Attacks Without an Accuracy Tradeoff

Nov 18, 2020
Figure 1 for Strong Data Augmentation Sanitizes Poisoning and Backdoor Attacks Without an Accuracy Tradeoff
Figure 2 for Strong Data Augmentation Sanitizes Poisoning and Backdoor Attacks Without an Accuracy Tradeoff
Figure 3 for Strong Data Augmentation Sanitizes Poisoning and Backdoor Attacks Without an Accuracy Tradeoff
Figure 4 for Strong Data Augmentation Sanitizes Poisoning and Backdoor Attacks Without an Accuracy Tradeoff
Viaarxiv icon

Towards Accurate Quantization and Pruning via Data-free Knowledge Transfer

Oct 14, 2020
Figure 1 for Towards Accurate Quantization and Pruning via Data-free Knowledge Transfer
Figure 2 for Towards Accurate Quantization and Pruning via Data-free Knowledge Transfer
Figure 3 for Towards Accurate Quantization and Pruning via Data-free Knowledge Transfer
Figure 4 for Towards Accurate Quantization and Pruning via Data-free Knowledge Transfer
Viaarxiv icon

Breaking certified defenses: Semantic adversarial examples with spoofed robustness certificates

Add code
Mar 19, 2020
Figure 1 for Breaking certified defenses: Semantic adversarial examples with spoofed robustness certificates
Figure 2 for Breaking certified defenses: Semantic adversarial examples with spoofed robustness certificates
Figure 3 for Breaking certified defenses: Semantic adversarial examples with spoofed robustness certificates
Figure 4 for Breaking certified defenses: Semantic adversarial examples with spoofed robustness certificates
Viaarxiv icon

Label Smoothing and Logit Squeezing: A Replacement for Adversarial Training?

Add code
Oct 25, 2019
Figure 1 for Label Smoothing and Logit Squeezing: A Replacement for Adversarial Training?
Figure 2 for Label Smoothing and Logit Squeezing: A Replacement for Adversarial Training?
Figure 3 for Label Smoothing and Logit Squeezing: A Replacement for Adversarial Training?
Figure 4 for Label Smoothing and Logit Squeezing: A Replacement for Adversarial Training?
Viaarxiv icon

Adversarially robust transfer learning

Add code
May 20, 2019
Figure 1 for Adversarially robust transfer learning
Figure 2 for Adversarially robust transfer learning
Figure 3 for Adversarially robust transfer learning
Figure 4 for Adversarially robust transfer learning
Viaarxiv icon

Adversarial Training for Free!

Add code
Apr 29, 2019
Figure 1 for Adversarial Training for Free!
Figure 2 for Adversarial Training for Free!
Figure 3 for Adversarial Training for Free!
Figure 4 for Adversarial Training for Free!
Viaarxiv icon