Landmarks-based Kernelized Subspace Alignment for Unsupervised Domain Adaptation

International audienceDomain adaptation (DA) has gained a lot of success in the recent years in Computer Vision to deal with situations where the learning process has to transfer knowledge from a source domain to a target domain. In this paper, we introduce a novel parameter free unsupervised DA approach based on both subspace alignment and the selection of landmarks similarly distributed between the two domains. Those landmarks are selected so as to reduce the discrepancy between the domains and then to allow a non linearly projection of the data in the same space where an efficient subspace alignment (in closed-form) is performed. We carry out a large experimental comparison in visual domain adaptation that shows that our new method outperforms the most recent DA approaches