Bidirectional One-Shot Unsupervised Domain Mapping

TL;DR

This paper introduces a novel bidirectional one-shot unsupervised domain mapping method that effectively translates between a richly sampled domain and a single-sample domain without weight sharing, outperforming existing solutions.

Contribution

The method enables bidirectional domain mapping between a single sample and a rich dataset using separate encoders and decoders, without weight sharing, and demonstrates superior performance.

Findings

Effective mapping from single sample to rich domain and vice versa

Outperforms existing solutions in one-shot domain translation

Code is publicly available for reproducibility

Abstract

We study the problem of mapping between a domain $A$, in which there is a single training sample and a domain $B$, for which we have a richer training set. The method we present is able to perform this mapping in both directions. For example, we can transfer all MNIST images to the visual domain captured by a single SVHN image and transform the SVHN image to the domain of the MNIST images. Our method is based on employing one encoder and one decoder for each domain, without utilizing weight sharing. The autoencoder of the single sample domain is trained to match both this sample and the latent space of domain $B$. Our results demonstrate convincing mapping between domains, where either the source or the target domain are defined by a single sample, far surpassing existing solutions. Our code is made publicly available at https://github.com/tomercohen11/BiOST