Lecturer: Dr. Wu xiang, The school of Engineering and Information Technology , UTS　

　　Time:　9:00 AM, July 29, 2013

　　Place: The Auditorium room, The school of Logistics Engineering

　　Title:

　　1.Vision and Learning - Research Capability Overview on UTS Team

　　Abstract: Through this talk, a comprehensive overview will be given on recent research outcomes from our teams. The research topics span over computer vision, image processing, pattern recognition, machine learning, and multimedia. The relevant demo video will be presented to demonstrate the performance achieved in various computer vision applications such as human detection, car licence plate detection, people counting, cheque fraud detection and also in several multimedia applications such as sport event detection and image categorisation. While showing the outcomes, the relevant algorithm/approach framework will be introduced, which covers several key research components widely seen in this area like feature extraction, feature selection, supervised training, regression, and sparse representation.

　　2.View-invariant Gait Recognition based on Low-rank Textures despite Gross Sparse Errors

　　Abstract: The study of human gait is innate to human interest and pervades many fields, including biometrics, clinical analysis, computer animation, and robotics. From a surveillance perspective, gait recognition is an attractive modality because it may be performed surreptitiously and at a distance in an unconstrained environment, which is not possible with other biometric techniques such as face recognition. Meanwhile, gait analysis plays an important role in human computer interaction and entertainment, such as computer games and character animation. In a real-world environment, there are various factors significantly affecting human gait including people dressing in different clothes, walking while carrying different objects, on different surfaces, in different shoes, at variable speeds and from arbitrary views. These are the main reasons constraining gait recognition being applied to real applications. View angle change has been regarded as one of the biggest challenges and unresolved problem to gait recognition as it can change significantly the available visual features for matching. In this talk, I will introduce our outcomes on the novel solutions to tackle the challenges of cross-view gait recognition. The relevant outcomes have been published on several IEEE transactions and top research conferences.