Open Set Domain Adaptation by Backpropagation

Abstract

Numerous algorithms have been proposed for transferring knowledge from a label-rich domain (source) to a label-scarce domain (target). Almost all of them are proposed for closed-set scenario, where the source and the target domain completely share the class of their samples. We call the shared class the "known class." However, in practice, when samples in target domain are not labeled, we cannot know whether the domains share the class. A target domain can contain samples of classes that are not shared by the source domain. We call such classes the "unknown class" and algorithms that work well in the open set situation are very practical. However, most existing distribution matching methods for domain adaptation do not work well in this setting because unknown target samples should not be aligned with the source. In this paper, we propose a method for an open set domain adaptation scenario which utilizes adversarial training. A classifier is trained to make a boundary between the source and the target samples whereas a generator is trained to make target samples far from the boundary. Thus, we assign two options to the feature generator: aligning them with source known samples or rejecting them as unknown target samples. This approach allows to extract features that separate unknown target samples from known target samples. Our method was extensively evaluated on domain adaptation setting and outperformed other methods with a large margin in most setting.

Invariance Matters: Exemplar Memory for Domain Adaptive Person Re-Identification

“…To tackle this problem, Busto and Grall [3] develop a method to learn a mapping from the source domain to the target domain by discarding unknown-class target samples. Recently, an adversarial learning framework [22] is proposed to separate target samples into known and unknown classes, and reject unknown classes during feature alignment. In this paper, we study the problem of UDA in person re-ID, where the classes are totally different between the source and target domains.…”

Learning to Adapt Invariance in Memory for Person Re-identification

“…To tackle this problem, Busto and Grall [4] develop a method to learn a mapping from the source domain to the target domain by jointly predicting unknownclass target samples and discarding them. Saito et al [35] introduce an adversarial learning framework to separate target samples into known and unknown classes. Meanwhile, unknown classes are rejected during feature alignment.…”

“…A popular approach is to align the feature distributions of both domains, but it does not readily apply to the context of re-ID. Since domain adaptation in re-ID is a special open-set problem [4], [35], [38], where the source and target domains have completely disjoint classes/identities. For such label constraint, directly aligning the feature distributions of two domains will align the samples from different classes and may be detrimental to the adaptation accuracy.…”

VisDA: A Synthetic-to-Real Benchmark for Visual Domain Adaptation

“…Open-set SVM [49] rejects unknown samples lower than a pre-defined probability threshold, and works well when known and unknown samples belong to the same domain. Recent AODA [51] method utilizes adversarial approach for both domain adaptation and unknown outlier detection.…”