Discriminator-free Unsupervised Domain Adaptation for Multi-label Image Classification

TL;DR

This paper introduces a discriminator-free adversarial approach for unsupervised domain adaptation in multi-label image classification, improving performance by directly leveraging task-specific classifiers and Gaussian mixture models.

Contribution

The paper proposes a novel discriminator-free adversarial method using a Gaussian Mixture Model-based critic derived from the classifier for better domain adaptation in MLIC.

Findings

Outperforms state-of-the-art methods in precision.

Requires fewer parameters than existing approaches.

Effective across various domain shifts.

Abstract

In this paper, a discriminator-free adversarial-based Unsupervised Domain Adaptation (UDA) for Multi-Label Image Classification (MLIC) referred to as DDA-MLIC is proposed. Recently, some attempts have been made for introducing adversarial-based UDA methods in the context of MLIC. However, these methods which rely on an additional discriminator subnet present one major shortcoming. The learning of domain-invariant features may harm their task-specific discriminative power, since the classification and discrimination tasks are decoupled. Herein, we propose to overcome this issue by introducing a novel adversarial critic that is directly deduced from the task-specific classifier. Specifically, a two-component Gaussian Mixture Model (GMM) is fitted on the source and target predictions in order to distinguish between two clusters. This allows extracting a Gaussian distribution for each component. The resulting Gaussian distributions are then used for formulating an adversarial loss based on a Frechet distance. The proposed method is evaluated on several multi-label image datasets covering three different types of domain shift. The obtained results demonstrate that DDA-MLIC outperforms existing state-of-the-art methods in terms of precision while requiring a lower number of parameters. The code is publicly available at github.com/cvi2snt/DDA-MLIC.