One-on-one tutoring is a critical component of teaching and learning. This project investigates technology to deploy an anywhere-tutor anywhere-student system that permits a tutor to provide instruction via an animated avatar. Exploiting psycholinguistic information, the system translates time-situated pointing and communicative gestures performed on a LCD tablet in conjunction with speech into a stream of behaviorally correct, spatially situated 3D gestures performed by the avatar tutor. A camera tracks the student, facilitating socially aware behavior by the avatar. The system exploits the student's ability to use body motion cues in the embodied avatar to direct her attention to appropriate locations of projected graphics that serve as artifacts of instruction. The technology will potentially facilitate a tutoring program in which students may connect to a pool of tutors for help with a variety of topics. The approach essentially implements tutoring 'telephone handsets' at both ends of the interaction. Tutors may be distributed across multiple sites at their convenience and take calls as they are needed. Widespread access to such tutoring will facilitate the vision of universally available educational opportunities. The project will investigate the necessary artificial intelligence and interaction technologies and evaluate the efficacy of this tutoring methodology.
