IBM Exploratory Computer Vision Group - IBM's Computer Vision Group - Research areas include PeopleVision, Biometrics and controlling computers by observing the user - http://researchweb.watson.ibm.com/ecvg/
Active Vision Lab, University of Oxford - Currently(2004) working in Whole body motion, hand tracking, wearable robot, Simultaneous Localisation And Mapping (SLAM) with a single camera. - http://www.robots.ox.ac.uk/~lav/
Robust Image Understanding Laboratory - Computer Vision Lab at Rutgers University, NJ. Research topics include Hetereoscedasticity, Robust Analysis, Bootstrap, Retrieval. - http://www.caip.rutgers.edu/riul/index.html
Computer Vision Lab at University of Central Florida - Has a large section of downloadable papers. Section on Research Experience for Undergraduates (REU) includes a sample proposal and papers co-authored by REU students - http://www.cs.ucf.edu/~vision
Face Expression Recognition Project at UPM - Computer Vision-based face expression recognition research project. We are doing robust face tracking, background maintenance, colour constancy based tracking, face expression recognition, or camera calibration. - http://www.dia.fi.upm.es/~lbaumela/FaceExpressionRecognition
Machine Vision Group at University of Ljubljana, Slovenia - Part of the Laboratory of Artificial Perception, this page includes biographies of group members and their research work, including list of publications - http://vision.fe.uni-lj.si
UMASS Computer Vision Laboratory - Researching and investigating scientific principles underlying the construction of integrated vision systems. - http://vis-www.cs.umass.edu/
U.C. Berkeley Computer Vision Group - Contains description of research projects on segmentation, shape, object recognition, statistics of natural images, and human motion analysis. Publications, software and course materials are available for download. - http://www.cs.berkeley.edu/projects/vision/vision_group.html
Automatic Gait Recognition at Southampton - Describe new approaches to recognising people by the way they walk. This is a new biometric, with particular application advantages (availability at distance, and non contact). Early results show that by computer vision we can indeed recognise people by t - http://www.gait.ecs.soton.ac.uk/