Abstract
This paper presents a novel unsupervised domain adaptation method for cross-domain visual recognition. We propose a unifified framework that reduces the shift between domains both statistically and geometrically, referred to as Joint Geometrical and Statistical Alignment (JGSA). Specififically, we learn two coupled projections that project the source domain and target domain data into lowdimensional subspaces where the geometrical shift and distribution shift are reduced simultaneously. The objective function can be solved effificiently in a closed form. Extensive experiments have verifified that the proposed method signififi- cantly outperforms several state-of-the-art domain adaptation methods on a synthetic dataset and three different real world cross-domain visual recognition tasks.