Multiclass domain adaptation with iterative manifold alignment

摘要: We propose a novel approach for multiclass domain adaptation using an iterative manifold alignment technique inspired by the TRiplet-based Iterative ALignment (TRIAL) protein structure algorithm. Our learns rigid transformation each class set of automatically-selected pivot samples that characterize relative relationships between classes in two similar, but not identical, feature spaces. demonstrate our robustly reconciles domain-specific differences similar hyperspectral images captured under different conditions, and yields more accurate results than recently-proposed techniques. evaluate method on pair real-world Cuprite, NV, provide MATLAB implementation algorithm, available online.