Our research group performs research in several areas of computer vision , robotics and medical imaging. In computer vision, we design methods for dynamic vision (tracking, visual servoing, on-line 3D modeling from images, dynamic textures). In medical imaging, we look at applications of variational methods and PDEs for various tasks like segmentation, registration, tumor growth prediction on MRI and DTI (tensor) data. In robotics, we design and build vision-based software for autonomous and semi-autonomous control of manipulators and mobile manipulators. We focus on applications for unstructured environments.