Face recognition based on Riemannian manifold learning

In recent years, manifold learning becomes a hot study topic in the field of artificial intelligence and pattern recognition, and it is a nonlinear dimensionality reduction technique. This paper presents a novel and efficient manifold learning, called Riemannian manifold learning (RML), which is efficient for many nonlinear dimensionality reduction problems. The nonlinear dimensionality reduction problems are solved by constructing Riemannian normal coordinates. RML is applied to face recognition. The experimental results on the open human face databases have demonstrated that our RML algorithm has its effectiveness. Compared to some classical manifold learning algorithms, such as LLE and ISOMAP, the face recognition accuracy of RML algorithm is higher than that of them.