报告题目：Light Field Vision for Transparent Object Categorization and Segmentation [Slides]
报告摘要：Recognizing the object category and detecting a certain object in the image are two important object recognition tasks, but previous appearance-based methods cannot deal with the transparent objects since the appearance of a transparent object dramatically changes when the background varies. Our proposed methods overcome previous problems using the novel features extracted from a light-field image. We propose a light field distortion (LFD) feature, which is background-invariant, for transparent object recognition. Light field linearity (LF-linearity) is proposed to measure the likelihood of a point comes from the transparent object or not. The occlusion detector is designed to locate the occlusion boundary in the light field image. Transparent object categorization is performed by incorporating the LFD feature into the bag-of-features approach for recognizing the category of transparent object. Transparent object segmentation is realized by solving the pixel labeling problem. An energy function is defined and Graph-cut algorithm is applied for optimizing the pixel labeling problem. The regional term and boundary term are from the LF-linearity and occlusion detector output. Light field datasets (available by request) are acquired for the transparent object categorization and segmentation. The results demonstrate that the proposed methods successfully categorize and segment transparent objects from a light field image.
 Yichao Xu, Hajime Nagahara, Atsushi Shimada, Rin-ichiro Taniguchi, “TransCut: Transparent Object Segmentation from a Light-Field Image”, ICCV 2015, Santiago, Chile
 Yichao Xu, Kazuki Maeno, Hajime Nagahara, Atsushi Shimada, Rin-ichiro Taniguchi. “Light Field Distortion Feature for Transparent Object Classification”. Computer Vision and Image Understanding (CVIU), Vol.139, pp.122-135, 2015
Yichao Xu is a Postdoc researcher of Laboratory for Image and Media Understanding (LIMU), Kyushu University, Fukuoka, Japan, where he received his Ph.D. degrees in September 2015. His research interests are computer vision and computational photography. He has served as a reviewer for top computer vision conferences CVPR/ICCV/ECCV. Prior to Kyushu University, Yichao received his Master degree from University of Chinese Academy of Sciences, China in 2010, and his Bachelor degree from Beijing Electronic Science and Technology Institute, China in 2007. For more details please visit: http://xuyichao.cn
报告嘉宾2：Huiyu Zhou (The Queen’s University of Belfast, UK)
报告题目：Computer vision techniques for video surveillance [Slides]
报告摘要：Experienced security analysts often profile individuals in the scene to determine their threat. An expert can identify individuals who look as though they may cause trouble before any anti-social behaviour has occurred. Therefore, a key to automatic threat assessment is to be able to automatically profile people in the scene based on their gender and age. In this talk, I am going to introduce the challenges and promising solutions in my recent studies of age and gender classification. Afterwards, I will summarise the recent research progress in human behaviour analysis. In particular, I will talk about (1) The techniques that I helped to develop in order to automatically extract and select features from face images for robust age recognition, (2) An effective combination of facial and full body measurements for gender classification, (3) Human tracking and trajectory clustering approaches to handle challenging circumstances such as occlusions and pose variations, and (4) event reasoning in smart transport video surveillance.
 Hong, X., Huang, Y., Ma, W., Varadarajan, S., Miller, P., Liu, W., Romero, M.J., del Rincon, J.M. and Zhou, H. “Evidential event inference in transport video surveillance”. Computer Vision and Image Understanding, accepted.
 Ma, J., Liu, W., Miller, P. and Zhou, H. “An evidential fusion approach for gender profiling”. Information Sciences, Vol. 333, March, 10-20, 2016.
 Xin Hong, Wenjun Ma, Yan Huang, Paul C. Miller, Weiru Liu, Huiyu Zhou: Evidence Reasoning for Event Inference in Smart Transport Video Surveillance. ICDSC 2014: 36:1-36:6.
 Huiyu Zhou, Hushing Hu, Honghai Liu, Jinshan Tang: Classification of Upper Limb Motion Trajectories Using Shape Features. IEEE Transactions on Systems, Man, and Cybernetics, Part C 42(6): 970-982 (2012).
 Huiyu Zhou, Paul C. Miller, Jianguo Zhang: Age classification using Radon transform and entropy based scaling SVM. BMVC 2011: 1-12.
 Huiyu Zhou, Yuan Yuan, Chunmei Shi: Object tracking using SIFT features and mean shift. Computer Vision and Image Understanding 113(3): 345-352 (2009).
 Huiyu Zhou, Murtaza Taj, Andrea Cavallari: Target Detection and Tracking With Heterogeneous Sensors. IEEE J. Sel. Topics Signal Processing 2(4): 503-513 (2008).
Dr. Zhou’s research specialisation is in machine learning on image analysis. This is an interdisciplinary research area that draws his experience and training onto image understanding, robotics/intelligent systems and data mining. Image analysis has been continuously applied in the fields as diverse as visual surveillance, medicine, robotics and human-computer interface. The aim is to develop algorithms/systems that enable effective and efficient image segmentation, object detection and tracking, 3-D reconstruction and event search and retrieval. His research to date has focused on the development of machine learning algorithms that allows a system to better describe images while representing/inferring structures.
Huiyu Zhou obtained a Bachelor of Engineering degree in Radio Technology from the Huazhong University of Science and Technology of China, and a Master of Science degree in Biomedical Engineering from the University of Dundee of United Kingdom, respectively. He was then awarded a Doctor of Philosophy degree in Computer Vision from the Heriot-Watt University, Edinburgh, United Kingdom. Dr. Zhou is presently a lecturer at the Queen’s University of Belfast. He has published over 120 peer reviewed papers in the field. He was the recipient of “CVIU 2012 Most Cited Paper Award” and was shortlisted for “MBEC 2006 Nightingale Prize”. He also won one of the “Best Paper Awards” in the 1993 Annual Conference of China Association for Medical Devices Industry. He currently serves as the Editor-in-Chief of Recent Advances in Electrical & Electronic Engineering, and is on the Editorial Boards of several refereed journals. He is a Guest Co-Editor of Pattern Recognition, Neurocomputing, Signal Processing, Signal, Image and Video Processing, Scientific World Journal and Journal of Electrical and Computer Engineering, and a Co-Chair of International Workshop on Sparse Representation for Event Detection in Multimedia (SRED’11). He serves or has served as a technical program committee for 300 conferences in signal and image processing and a reviewer for 90 refereed journals including 19 IEEE Transactions/Journals.