The goal of the MPLab is to gain insights into how the brain works by developing embodied systems that solve problems similar to those encountered by the brain. We focus on systems that perceive and interact with humans in real time using natural communication channels (e.g., visual, auditory, and tactile information). To this effect we are developing perceptual primitives to detect and track human faces and to recognize facial expressions. We are also developing probabilistic models for integrating multiple sensory modalities, and actions. Developing such systems requires a multidisciplinary approach that combines mathematical modeling, machine learning techniques, computational modeling of brain function, and behavioral experiments. Applications include personal robots, automatic tutoring systems, and automatic assessment of affective disorders.