Unsupervised Day-Night Domain Adaptation with a Physics Prior for Image Classification

Unsupervised Day-Night Domain Adaptation with a Physics Prior for Image Classification

Brouwer, Gees (TU Delft Electrical Engineering, Mathematics and Computer Science)

Lengyel, A. (mentor)
van Gemert, J.C. (mentor)
Marroquim, Ricardo (graduation committee)

Delft University of Technology

Computer Science and Engineering

CSE3000 Research Project

2022-01-28

While deep neural networks show great potential for being part of safety-critical applications such as autonomous driving, covering their sensitivity to illumination shifts by adding training data is of- ten non-trivial. The undesired illumination shift between train and test data can be addressed by domain adaptation methods. Recent work [9] has demonstrated performance improvements with a novel zero-shot domain adaptation setting by in- troducing a physics-based visual inductive prior - a trainable Color Invariant Convolution (CIConv) layer - aiming to transform its input to a more do- main invariant representation.
We compare the performance of image classifica- tion for day-night domain adaptation in the zero- shot and the unsupervised setting, and explore the effectiveness of using CIConv in both settings. We show that unsupervised domain adaptation reduces the day-night distribution shift similarly to CIConv in the zero-shot setting. We demonstrate improved performance when CIConv and unsupervised day- night domain adaptation are combined.

http://resolver.tudelft.nl/uuid:7ea0396b-1bfa-4ebe-b976-b8432ccbcdff

Student theses

bachelor thesis

© 2022 Gees Brouwer