Matching Embeddings for Domain Adaptation

TL;DR

This paper introduces AVDA, a semi-supervised domain adaptation method that leverages deep variational embeddings and adversarial training to improve classification across different domains, especially with limited labeled data.

Contribution

The paper proposes AVDA, a novel approach combining variational embeddings and adversarial methods for effective semi-supervised domain adaptation.

Findings

AVDA outperforms previous methods in semi-supervised, few-shot scenarios.

AVDA requires fewer labeled samples to achieve high accuracy.

The method effectively aligns source and target domain representations.

Abstract

In this work we address the problem of transferring knowledge obtained from a vast annotated source domain to a low labeled target domain. We propose Adversarial Variational Domain Adaptation (AVDA), a semi-supervised domain adaptation method based on deep variational embedded representations. We use approximate inference and domain adversarial methods to map samples from source and target domains into an aligned class-dependent embedding defined as a Gaussian Mixture Model. AVDA works as a classifier and considers a generative model that helps this classification. We used digits dataset for experimentation. Our results show that on a semi-supervised few-shot scenario our model outperforms previous methods in most of the adaptation tasks, even using a fewer number of labeled samples per class on target domain.