Sspeaker：Prof.Ji Hui(National University of Singapore)
Title: Sparse coding: Models, algorithms and applications
Abstract: In recent years, sparse coding has been widely used in many applications ranging from image processing to pattern recognition. The basic concept of sparse coding is to learn an over-complete dictionary from data over which the data can be sparsely approximation. In the first part of the talk, I will first present several sparse coding motivated methods for solving various classification and recognition problems, including face recognition, object recognition, texture classification, and classification tasks in neural sciences. All these models require solving a class of challenging non-smooth and non-convex optimization problems. In the second part of the talk, I will present an alternating iteration scheme for solving such problems, which is the first available one that guarantees the global convergence property, i.e. the whole sequence of iterates is convergent and converges to a critical point. The practical beneﬁt of the proposed method is validated in a wide range of applications in classification and recognition.