Descriptor Learning via Supervised Manifold Regularization for Multioutput Regression

摘要: Multioutput regression has recently shown great ability to solve challenging problems in both computer vision and medical image analysis. However, due the huge variability ambiguity, it is fundamentally handle highly complex input-target relationship of multioutput regression, especially with indiscriminate high-dimensional representations. In this paper, we propose a novel supervised descriptor learning (SDL) algorithm for which can establish discriminative compact feature representations improve multivariate estimation performance. The SDL formulated as generalized low-rank approximations matrices manifold regularization. able simultaneously extract features closely related targets remove irrelevant redundant information by transforming raw into new low-dimensional space aligned targets. achieved while largely reduces ambiguity enables more accurate efficient estimation. We conduct extensive evaluation proposed on synthetic data real-world tasks Experimental results have that achieve high accuracy all outperforms algorithms state arts. Our method establishes framework be widely used boost performance different applications.