Can I read Class-Incremental Domain Adaptation on EtoBox?

Class-Incremental Domain Adaptation by Kundu, Jogendra Nath; Venkatesh, Rahul Mysore; Venkat, Naveen; Revanur, Ambareesh; Babu, R. Venkatesh is a scholarly article available to read on EtoBox.

What is Class-Incremental Domain Adaptation about?

We introduce a practical Domain Adaptation (DA) paradigm called Class-Incremental Domain Adaptation (CIDA). Existing DA methods tackle domain-shift but are unsuitable for learning novel target-domain classes. Meanwhile, class-incremental (CI) methods enable learning of new classes in absence of source training data but fail under a domain-shift without labeled supervision. In this work, we effectively identify the limitations of these approaches in the CIDA paradigm. Motivated by theoretical and empirical observations, we propose an effective method, inspired by prototypical networks, that enables classification of target samples into both shared and novel (one-shot) target classes, even under a domain-shift. Our approach yields superior performance as compared to both DA and CI methods in the CIDA paradigm.

Author

Kundu, Jogendra Nath; Venkatesh, Rahul Mysore; Venkat, Naveen; Revanur, Ambareesh; Babu, R. Venkatesh

Published

2020

Language

EN