Deep neural networks are very successful on many vision tasks, but hard to interpret due to their black box nature. To overcome this, various post-hoc attribution methods have been proposed to identify image regions most influential to the models' …
Despite being highly performant, deep neural networks might base their decisions on features that spuriously correlate with the provided labels, thus hurting generalization. To mitigate this, ‘model guidance’ has recently gained popularity, i.e. the …
Studying How to Efficiently and Effectively Guide Models with Explanations
Towards Better Understanding Attribution Methods
Deep neural networks are very successful on many vision tasks, but hard to interpret due to their black box nature. To overcome this, various post-hoc attribution methods have been proposed to identify image regions most influential to the models' …
Deep neural networks have been shown to be susceptible to adversarial examples -- small, imperceptible changes constructed to cause mis-classification in otherwise highly accurate image classifiers. As a practical alternative, recent work proposed …
Adversarial Training against Location-Optimized Adversarial Patches
Research Internship at the University of Tokyo in the summer of 2018
Research Internship at Robert Bosch in the summer of 2017
Implementation of the model described in "Training deep neural nets to aggregate crowdsourced responses." by Gaunt et al.