3 papers
cs.CV2023
Frequency-Based Vulnerability Analysis of Deep Learning Models against Image Corruptions
Harshitha Machiraju, Michael H. Herzog, Pascal Frossard
Deep learning models often face challenges when handling real-world image corruptions. In response, researchers have developed image corruption datasets to evaluate the performance…
cs.CV2022
CLAD: A Contrastive Learning based Approach for Background Debiasing
Ke Wang, Harshitha Machiraju, Oh-Hyeon Choung +2
Convolutional neural networks (CNNs) have achieved superhuman performance in multiple vision tasks, especially image classification. However, unlike humans, CNNs leverage spurious…
q-bio.NC2022
A comment on Guo et al. [arXiv:2206.11228]
Ben Lonnqvist, Harshitha Machiraju, Michael H. Herzog
In a recent article, Guo et al. [arXiv:2206.11228] report that adversarially trained neural representations in deep networks may already be as robust as corresponding primate IT ne…