Self-ensembling for visual domain adaptation

TL;DR

This paper applies self-ensembling, based on the mean teacher method, to visual domain adaptation, achieving state-of-the-art results across multiple benchmarks and close to supervised accuracy in small datasets.

Contribution

It introduces modifications to the mean teacher approach specifically for challenging domain adaptation scenarios, demonstrating its effectiveness.

Findings

Achieves state-of-the-art results on VISDA-2017 benchmark.

Outperforms prior methods on small image benchmarks.

Close to supervised accuracy in small datasets.

Abstract

This paper explores the use of self-ensembling for visual domain adaptation problems. Our technique is derived from the mean teacher variant (Tarvainen et al., 2017) of temporal ensembling (Laine et al;, 2017), a technique that achieved state of the art results in the area of semi-supervised learning. We introduce a number of modifications to their approach for challenging domain adaptation scenarios and evaluate its effectiveness. Our approach achieves state of the art results in a variety of benchmarks, including our winning entry in the VISDA-2017 visual domain adaptation challenge. In small image benchmarks, our algorithm not only outperforms prior art, but can also achieve accuracy that is close to that of a classifier trained in a supervised fashion.