Preprint out for the paper: Universal Adversarial Perturbations: A Survey!

The new survey paper on the topic of Universal Adversarial Perturbations is out now on arxiv. The paper covers the basic terminologies and concepts that build up the subject matter of Adversarial Perturbations and then all the recent literature work to further develop the field. Lastly it also delves into the applications of the same and the future work directions possible in this field.

Checkout the preprint here.

Vision and Language Group
Vision and Language Group

Deep Learning Research Group