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

Add to Chrome

Add to Firefox

Picture for Akshay Mehra

Akshay Mehra

Understanding the Robustness of Multi-Exit Models under Common Corruptions

Dec 03, 2022
Akshay Mehra, Skyler Seto, Navdeep Jaitly, Barry-John Theobald

Figure 1 for Understanding the Robustness of Multi-Exit Models under Common Corruptions

Figure 2 for Understanding the Robustness of Multi-Exit Models under Common Corruptions

Figure 3 for Understanding the Robustness of Multi-Exit Models under Common Corruptions

Figure 4 for Understanding the Robustness of Multi-Exit Models under Common Corruptions

* 16 pages, 22 figures

Access Paper or Ask Questions

Certified Adversarial Defenses Meet Out-of-Distribution Corruptions: Benchmarking Robustness and Simple Baselines

Dec 01, 2021
Jiachen Sun, Akshay Mehra, Bhavya Kailkhura, Pin-Yu Chen, Dan Hendrycks, Jihun Hamm, Z. Morley Mao

Figure 1 for Certified Adversarial Defenses Meet Out-of-Distribution Corruptions: Benchmarking Robustness and Simple Baselines

Figure 2 for Certified Adversarial Defenses Meet Out-of-Distribution Corruptions: Benchmarking Robustness and Simple Baselines

Figure 3 for Certified Adversarial Defenses Meet Out-of-Distribution Corruptions: Benchmarking Robustness and Simple Baselines

Figure 4 for Certified Adversarial Defenses Meet Out-of-Distribution Corruptions: Benchmarking Robustness and Simple Baselines

* 21 pages, 15 figures, and 9 tables

Access Paper or Ask Questions

Machine Learning with Electronic Health Records is vulnerable to Backdoor Trigger Attacks

Jun 15, 2021
Byunggill Joe, Akshay Mehra, Insik Shin, Jihun Hamm

Figure 1 for Machine Learning with Electronic Health Records is vulnerable to Backdoor Trigger Attacks

Figure 2 for Machine Learning with Electronic Health Records is vulnerable to Backdoor Trigger Attacks

Figure 3 for Machine Learning with Electronic Health Records is vulnerable to Backdoor Trigger Attacks

Figure 4 for Machine Learning with Electronic Health Records is vulnerable to Backdoor Trigger Attacks

* AAAI 2021 Workshop on Trustworthy AI for Healthcare

Access Paper or Ask Questions

Penalty Method for Inversion-Free Deep Bilevel Optimization

Nov 08, 2019
Akshay Mehra, Jihun Hamm

Figure 1 for Penalty Method for Inversion-Free Deep Bilevel Optimization

Figure 2 for Penalty Method for Inversion-Free Deep Bilevel Optimization

Figure 3 for Penalty Method for Inversion-Free Deep Bilevel Optimization

Figure 4 for Penalty Method for Inversion-Free Deep Bilevel Optimization

* 16 Pages, 7 figures

Access Paper or Ask Questions

Machine vs Machine: Minimax-Optimal Defense Against Adversarial Examples

Jun 27, 2018
Jihun Hamm, Akshay Mehra

Figure 1 for Machine vs Machine: Minimax-Optimal Defense Against Adversarial Examples

Figure 2 for Machine vs Machine: Minimax-Optimal Defense Against Adversarial Examples

Figure 3 for Machine vs Machine: Minimax-Optimal Defense Against Adversarial Examples

Figure 4 for Machine vs Machine: Minimax-Optimal Defense Against Adversarial Examples

Access Paper or Ask Questions

Fast Interactive Image Retrieval using large-scale unlabeled data

Feb 12, 2018
Akshay Mehra, Jihun Hamm, Mikhail Belkin

Figure 1 for Fast Interactive Image Retrieval using large-scale unlabeled data

Figure 2 for Fast Interactive Image Retrieval using large-scale unlabeled data

Figure 3 for Fast Interactive Image Retrieval using large-scale unlabeled data

Figure 4 for Fast Interactive Image Retrieval using large-scale unlabeled data

* 15 Pages, Submitted to KDD 2018

Access Paper or Ask Questions