Unsupervised Visual Domain Adaptation Using Subspace Alignment

• Published in: 2013 IEEE International Conference on Computer Vision pp. 2960 - 2967
• Main Authors: Fernando, Basura, Habrard, Amaury, Sebban, Marc, Tuytelaars, Tinne
• Format: Conference Proceeding Journal Article
• Language: English
• Published: IEEE 01.12.2013
• Subjects: Adaptation; Algorithms; Alignment; Computer vision; Conferences; Context; Covariance matrices; domain adaptation; Eigenvalues and eigenfunctions; Exact solutions; Manifolds; Mathematical analysis; Mathematical models; object recognition; Principal component analysis; subspace alignment; Subspaces; Support vector machines; Vectors
• ISSN: 1550-5499
• DOI: 10.1109/ICCV.2013.368

In this paper, we introduce a new domain adaptation (DA) algorithm where the source and target domains are represented by subspaces described by eigenvectors. In this context, our method seeks a domain adaptation solution by learning a mapping function which aligns the source subspace with the target one. We show that the solution of the corresponding optimization problem can be obtained in a simple closed form, leading to an extremely fast algorithm. We use a theoretical result to tune the unique hyper parameter corresponding to the size of the subspaces. We run our method on various datasets and show that, despite its intrinsic simplicity, it outperforms state of the art DA methods.