Open Set Domain Adaptation using Optimal Transport

TL;DR

This paper introduces a two-step optimal transport method for open set domain adaptation, effectively rejecting unknown classes and addressing class and feature shifts, with demonstrated superior performance.

Contribution

The paper proposes a novel two-step optimal transport framework for open set domain adaptation, handling unknown classes and distribution shifts simultaneously.

Findings

Outperforms recent state-of-the-art methods.

Effective rejection of unknown classes.

Robust to label-shift and covariate shift.

Abstract

We present a 2-step optimal transport approach that performs a mapping from a source distribution to a target distribution. Here, the target has the particularity to present new classes not present in the source domain. The first step of the approach aims at rejecting the samples issued from these new classes using an optimal transport plan. The second step solves the target (class ratio) shift still as an optimal transport problem. We develop a dual approach to solve the optimization problem involved at each step and we prove that our results outperform recent state-of-the-art performances. We further apply the approach to the setting where the source and target distributions present both a label-shift and an increasing covariate (features) shift to show its robustness.