Domain Adaptation for Rare Classes Augmented with Synthetic Samples

Domain Adaptation for Rare Classes Augmented with Synthetic Samples

Das, Tuhin (TU Delft Electrical Engineering, Mathematics and Computer Science; TU Delft Intelligent Systems)

Bruintjes, R. (mentor)
Lengyel, A. (mentor)
van Gemert, J.C. (mentor)
Beery, S. (mentor)
Zarras, A. (graduation committee)

Delft University of Technology

Computer Science and Engineering

CSE3000 Research Project

2021-07-01

To alleviate lower classification performance on rare classes in imbalanced datasets, a possible solution is to augment the underrepresented classes with synthetic samples. Domain adaptation can be incorporated in a classifier to decrease the domain discrepancy between real and synthetic samples. While domain adaptation is generally applied on completely synthetic source domains and real target domains, we explore how domain adaptation can be applied when only a single rare class is augmented with simulated samples. As a testbed, we use a camera trap animal dataset with a rare deer class, which is augmented with synthetic deer samples. We adapt existing domain adaptation methods to two new methods for the single rare class setting: DeerDANN, based on the Domain-Adversarial Neural Network (DANN), and DeerCORAL, based on deep correlation alignment (Deep CORAL) architectures. Experiments show that DeerDANN has the highest improvement in deer classification accuracy of 24.0% versus 22.4% improvement of DeerCORAL when compared to the baseline. Further, both methods require fewer than 10k synthetic samples, as used by the baseline, to achieve these higher accuracies. DeerCORAL requires the least number of synthetic samples (2k deer), followed by DeerDANN (8k deer).

Domain Adaptation
Synthetic Data
Data Augmentation
Camera Trap Dataset

http://resolver.tudelft.nl/uuid:8dd15a01-e60d-4b94-860b-db3091995a6f

Student theses

bachelor thesis

© 2021 Tuhin Das