Open Set Domain Adaptation using Optimal Transport
This addresses domain adaptation challenges for machine learning systems when target data includes unknown classes, which is common in real-world applications.
The paper tackles the problem of open set domain adaptation where target domains contain new classes not present in the source domain, using a two-step optimal transport approach that first rejects samples from new classes and then addresses target class ratio shift, achieving state-of-the-art performance improvements.
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.