Kobus Barnard
Location: (Tucson Arizona)
Personal Research Web Page: http://vision.cs.arizona.edu/
Keywords: learning 3D stochastic geometric models, extracting biological form from image data, grammar representations, L-systems, image annotation, region labeling, color constancy, shadow detection, educational video
Posted on: Tuesday, June 2nd, 2009
Broad Research Area: AI / Machine Learning / Robotics / Vision, Scientific/Medical Informatics
Research Interests:
Broadly defined, my research interests are extracting semantics from data, focusing on useful and parsimonious representations, using statistical learning approaches. Application areas of interest include: 1) general object recognition and image understanding; 2) learning and fitting models of biological form; 3) the application of computer vision to the organization and effective use of large image and video collections such as educational video; and 4) physics based vision problems such as understanding scene illumination.
