I am the director of a research group in Computer Vision. The researchers in my group address fundamental problems that enable complex systems with vision capabilities through which they can learn about, interact with, and take decisions about the environment. We focus on developing novel techniques to analyze image and video that provide a semantic understanding of the materials of a scene, the objects that are present, their actions, and interactions. The outcome of this research is directly applicable to areas such as video surveillance, multimedia applications for Internet and mobile systems, medical applications, biometrics, autonomous navigation, robotics, and many others.