Detach and Adapt: Learning Cross-Domain Disentangled Deep Representation

TL;DR

This paper introduces CDRD, a deep learning model that achieves cross-domain feature disentanglement and adaptation by leveraging annotated source data and unlabeled target data, improving unsupervised domain classification.

Contribution

The novel CDRD model enables joint cross-domain disentanglement and adaptation without requiring annotations in the target domain.

Findings

Qualitative verification of disentanglement capability

Effective for unsupervised domain classification tasks

Outperforms state-of-the-art image translation methods

Abstract

While representation learning aims to derive interpretable features for describing visual data, representation disentanglement further results in such features so that particular image attributes can be identified and manipulated. However, one cannot easily address this task without observing ground truth annotation for the training data. To address this problem, we propose a novel deep learning model of Cross-Domain Representation Disentangler (CDRD). By observing fully annotated source-domain data and unlabeled target-domain data of interest, our model bridges the information across data domains and transfers the attribute information accordingly. Thus, cross-domain joint feature disentanglement and adaptation can be jointly performed. In the experiments, we provide qualitative results to verify our disentanglement capability. Moreover, we further confirm that our model can be applied for solving classification tasks of unsupervised domain adaptation, and performs favorably against state-of-the-art image disentanglement and translation methods.