Research focus
Intelligent Visual Computing (IVC)
tnt group
Visual computing is an entire field of acquiring, analyzing and synthesizing visual data (such as digital images, videos and visual simulations) by means of computers. It requires profound and interdisciplinary scientific knowledge, in particular mathematics, physics, engineering, and cognitive science. The Intelligent Visual Computing (IVC) subgroup studies how to make computers efficiently perceive, process, and understand visual data and the ultimate goal is for computers to emulate the striking perceptual capability of human eyes and brains. Especially, it aims to propose and design general (or domain-oriented) technological, mathematical, and theoretical aspects of algorithms and methods, spanning most traditional but important research topics in machine learning, pattern analysis and data mining, as well as exploring the applicability of new-emerging theory and methods in related fields. Research results from the IVC subgroup are specifically expected to have fundamental contributions to several applications, namely biometric recognition, video surveillance, and cyber data understanding and security.
Intelligent Video Surveillance (IVS)
tnt group
Security has become a major world-wide concern since the event of September 11, 2001 in USA and the bomb attacks in London on the July 2005. Video surveillance is a critical component of any effective security system. Most current video surveillance systems are monitored by relatively small teams of human operators even though there may be a very large number of cameras. Typically a human watches a set of screens which cycle from one camera to another every few seconds. In addition to problems of fatigue and boredom, the human attention span is limited both spatially and temporally. To overcome these limitations, Intelligent Video Surveillance system , from the computer vision and pattern recognition field, is developed for the real-time monitoring of humans and vehicles. These systems can interpret the events in the camera and generate an alarm if a suspicious person or an abnormal activity is detected.
Biometric Acquisition and Recognition (BAR)
tnt group

The goal of research in BAR is to develop advanced biometric sensors and innovative recognition algorithms for accurate, fast and user-friendly personal identification in unconstrained environments. We have started biometrics research since 1998 and contributed greatly to iris, face (both 2D and 3D), fingerprint, palmprint, palm vein, handwriting, gait recognition and multi-biometrics areas. We have constructed a number of large-scale biometric databases and share them to public domain free of charge. Our current research aims to develop a long-range biometric sensing and recognition system for identifying multiple subjects in motion based on fusion of iris, face and palmprint features.

Cyber Data Understanding and Security (CDUS)
tnt group

We are now facing a more complicated digital world. We communicate via email, mobile phones; we watch online videos, play online games; we rely on GPS to get the travel destination; we also shop online by using credit cardsIt seems that everything relies on computers and networks now.

Cyber data understanding and security involves information understanding and protecting by preventing, detecting, and responding to potential attacks. By using data mining, machine learning and pattern recognition we do forensics investigation for information security.