New Approaches to Unsupervised Domain Adaptation
The cost of large scale data collection and annotation often makes the application of machine learning algorithms to new tasks or datasets prohibitively expensive. One approach circumventing this cost is training models on synthetic data where annotations are provided automatically. Despite their appeal, such models often fail to generalize from synthetic to real images, necessitating domain adaptation algorithms to manipulate these models before they can be successfully applied. We present two approaches to the problem of unsupervised visual domain adaptation. Our first approach, presented as a paper at NIPS 2016, focuses on learning a shared representation between the two domains. We show that explicitly modeling what is unique to each domain can improve a model's ability to extract domain-invariant features. Inspired by work on private-shared component analysis, we explicitly learn to extract image representations that are partitioned into two subspaces: one component which is private to each domain and one which is shared across domains. In our second approach, we propose a new model that learns, in an unsupervised manner, a transformation in the pixel space from one domain to the other. Our generative adversarial network (GAN)-based method adapts synthetic images to make them appear more realistic. Both methods outperform the state-of-the-art for unsupervised visual domain adaptation.
Dilip is a Research Scientist at Google’s Cambridge office (Massachusetts). His research interests are in Machine/Deep Learning for Computer Vision and Scientific Computing. From August 2013 to November 2014, he was a Postdoctoral Associate with Bill Freeman at MIT’s CSAIL Lab. In June 2013, he received his PhD from the Computer Science department at New York University , under the supervision of Rob Fergus. Dilip was awarded a Microsoft Research PhD Fellowship for 2010-2011, a Dean’s Dissertation Fellowship for 2012-2013 from the Graduate School of Arts and Sciences and the Janet Fabri prize (2013-2014) for outstanding dissertation in Computer Science.