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.

 时间地点：2017年12月19号上午9:00-10:00，工商楼二楼200-9.

欢迎广大师生参加！

李松