Real-time reading of American Sign Language (with Thad Starner doing
the signing).
Starner and Pentland[20] explored the extremes of gesture-based mapping by combining th pfinder statistics representation with hidden Markov modeling to interpret a forty word subset of American Sign Language (ASL). Using this approach they were able to produce a real-time ASL interpreter with a 99% sign recognition accuracy. Thad Starner is shown using this system in Fig. 5.4.
