报告题目: Multi-view Discriminant Analysis
 Meina Kan, Shiguang Shan,Haihong Zhang, Shihong Lao, and Xilin Chen. Multi-view Discriminant Analysis. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2015.
 Meina Kan, Shiguang Shan, Haihong Zhang, Shihong Lao, Xilin Chen. Multi-view Discriminant Analysis. European Conference on Computer Vision (ECCV), 2012.
摘要：In many computer vision systems, the same object can be observed at varying viewpoints or even by different sensors, which brings in the challenging demand for recognizing objects from distinct even heterogeneous views. In this talk, we will introduce our Multi-view Discriminant Analysis (MvDA) approach, which seeks for a single discriminant common space for multiple views in a non-pairwise manner by jointly learning multiple view-specific linear transforms. Specifically, our MvDA is formulated to jointly solve the multiple linear transforms by optimizing a generalized Rayleigh quotient, i.e., maximizing the between-class variations and minimizing the within-class variations from both intra-view and inter-view in the common space. By reformulating this problem as a ratio trace problem, the multiple linear transforms are achieved analytically and simultaneously through generalized eigenvalue decomposition.
报告人介绍：阚美娜，博士，毕业于中国科学院计算所，现为计算所副研究员。2014年获得CCF优秀博士学位论文奖以及中科院优秀博士学位论文奖。研究 领域为计算机视觉与模式识别，主要关注人脸识别、多视学习、半监督学习、迁移学习、深度学习等问题，相关成果已发表在TPAMI、IJCV、 TIP、CVPR、ICCV等相关领域主流国际期刊与会议上面。目前担任TPAMI、IJCV、TIP、TMM、TSMC、TNN等多个刊物的审稿人。
报告题目：From Dictionary of Visual Words to Subspaces: Locality-constrained Affine Subspace Coding
文章信息：Peihua Li, Xiaoxiao Lu, Qilong Wang. From Dictionary of Visual Words to Subspaces: Locality-constrained Affine Subspace Coding. Int. Conf. on Computer Vision and Pattern Recognition (CVPR), 2015, pp. 2348-2357, 2015.
报告摘要：The locality-constrained linear coding (LLC) is a very successful feature coding method in image classification. It makes known the importance of locality constraint which brings high efficiency and local smoothness of the codes. However, in the LLC method the geometry of feature space is described by an ensemble of representative points (visual words) while discarding the geometric structure immediately surrounding them. Such a dictionary only provides a crude, piecewise constant approximation of the data manifold. To approach this problem, we propose a novel feature coding method called locality-constrained affine subspace coding (LASC). The data manifold in LASC is characterized by an ensemble of subspaces attached to the representative points (or affine subspaces), which can provide a piecewise linear approximation of the manifold. Given an input descriptor, we find its top-k neighboring subspaces, in which the descriptor is linearly decomposed and weighted to form the first-order LASC vector. Inspired by the success of usage of higher-order information in image classification, we propose the second-order LASC vector based on the Fisher information metric for further performance improvement. We compare with state-of-the-art methods and experiments have shown the LASC method is very competitive.
报告人简介：Peihua Li is a professor of School of Information and Communication Engineering, Dalian University of Technology (DUT). Before that, he was a professor, an associate professor of School of Computer Science in Heilongjiang University. He received Ph.D degree in computer science and technology from Harbin Institute of Technology (HIT) in 2003, and then working for one year as a postdoctoral fellow at INRIA/IRISA, Rennes, France. He achieved the best Ph.D dissertation award from HIT in 2004, and honorary nomination of National Excellent Doctoral dissertation in China in 2005. He was supported by Program for New Century Excellent Talents in University of Chinese Ministry of Education. His team achieved the second-best ranking position among all the teams in ALISC 2015: Alibaba Large-scale Image Search Challenge. His research interests include computer vision, pattern recognition and statistical learning. He has published scientific papers in top journals and conferences including ICCV、CVPR and ECCV.