FDA: Fourier Domain Adaptation for Semantic Segmentation

TL;DR

This paper introduces a simple Fourier-based method for unsupervised domain adaptation in semantic segmentation, effectively reducing domain discrepancy without complex training procedures.

Contribution

The proposed Fourier Domain Adaptation method is a straightforward, training-free approach that achieves state-of-the-art results in semantic segmentation domain adaptation.

Findings

Achieves state-of-the-art performance on benchmarks.

Does not require adversarial training or complex optimization.

Effective in reducing domain discrepancy with simple Fourier transforms.

Abstract

We describe a simple method for unsupervised domain adaptation, whereby the discrepancy between the source and target distributions is reduced by swapping the low-frequency spectrum of one with the other. We illustrate the method in semantic segmentation, where densely annotated images are aplenty in one domain (synthetic data), but difficult to obtain in another (real images). Current state-of-the-art methods are complex, some requiring adversarial optimization to render the backbone of a neural network invariant to the discrete domain selection variable. Our method does not require any training to perform the domain alignment, just a simple Fourier Transform and its inverse. Despite its simplicity, it achieves state-of-the-art performance in the current benchmarks, when integrated into a relatively standard semantic segmentation model. Our results indicate that even simple procedures can discount nuisance variability in the data that more sophisticated methods struggle to learn away.