Dolltalk Cati Vaucelle and Hiroshi Ishii

Dolltalk is a gesture based storytelling system. Dolltalk captures, analyzes and interprets a set of gestures in parallel with analyzing changes in voice prosody. Using sensors and audio analysis, the system interprets the narrative structure of a story. The primary goal of Dolltalk is to invite children to discover narrative perspectives during storytelling play.
In our next research Picture This, we integrate the visual perspective that complements the storytelling in Dolltalk with how a child can produce movies from her toy environment as she plays.