DLDL: Dynamic label dictionary learning via hypergraph regularization

For classification tasks, dictionary learning based methods have attracted lots of attention in recent years. One popular way to achieve this purpose is introduce label information generate a discriminative represent samples. However, compared with traditional learning, category only achieves significant improvements supervised and has little positive influence on semi-supervised or unsupervised learning. To tackle issue, we propose Dynamic Label Dictionary Learning (DLDL) algorithm the soft matrix for unlabeled data. Specifically, employ hypergraph manifold regularization keep relations among original data, transformed labels consistent. We demonstrate efficiency proposed DLDL approach two kinds including remote sensing image human activity recognition. The demonstrated our method.