My primary research interests lie in visual scene understanding. I am mainly concerned with solving problems arising in visual perception and action recognition, which are key components in many vision applications such as surveillance, robot guidance, multi-modal image interpretation, remote sensing, and visual field monitoring. I also focus my attention on the processing of optical/sonar images for underwater vision applications. Recently I’ve had a chance to explore many aspects falling at the edges between computer vision, computer graphics, and machine learning, such as performance-based animation, procedural animation, shape descriptors and spectral geometry for shape analysis.