Generate To Adapt: Aligning Domains using Generative Adversarial Networks
Authors:
Preprint
Abstract
Domain Adaptation is an actively researched problem in Computer Vision. In this work, we propose an approach that leverages unsupervised data to bring the source and target distributions closer in a learned joint feature space. We accomplish this by inducing a symbiotic relationship between the learned embedding and a generative adversarial network. This is in contrast to methods which use the adversarial framework for realistic data generation and retraining deep models with such data. We demonstrate the strength and generality of our approach by performing experiments on three different tasks with varying levels of difficulty: (1) Digit classification (MNIST, SVHN and USPS datasets) (2) Object recognition using OFFICE dataset and (3) Domain adaptation from synthetic to real data. Our method achieves state-of-the art performance in most experimental settings and by far the only GAN-based method that has been shown to work well across different datasets such as OFFICE and DIGITS.Keywords
All Related Versions
This publication has 8 references indexed in Scilit:
- Synthetic to Real Adaptation with Generative Correlation Alignment Networks2018
- Unsupervised Pixel-Level Domain Adaptation with Generative Adversarial Networks2017
- Adversarial Discriminative Domain Adaptation2017
- Learning Deep Object Detectors from 3D Models2015
- Domain Adaptation via Transfer Component AnalysisIEEE Transactions on Neural Networks, 2010
- The Pascal Visual Object Classes (VOC) ChallengeInternational Journal of Computer Vision, 2009
- Gradient-based learning applied to document recognitionProceedings of the IEEE, 1998
- A database for handwritten text recognition researchIEEE Transactions on Pattern Analysis and Machine Intelligence, 1994