MouthType: text entry by hand and mouth

In this paper we describe a novel text entry method which uses coordinated motor action of hand and mouth. A vision based algorithm is used to gauge shape parameters of the cavity of the open mouth. These are mapped to a discrete set of input states which are combined with keypad input in a factorial manner to allow unambiguous input of a large number of symbols. The method is implemented and tested for an alphabetic writing system (the Roman alphabet) and a syllabic writing system (Japanese hiragana). We report the results of preliminary experiments to measure text entry speed and error rate. Categories & Subject Descriptors: H.5.2. User Interfaces: Input devices and strategies. General Terms: Human Factors

• Type: Conference
• Year: 2004
• Where: CHI
• Authors: Michael J. Lyons, Chi-Ho Chan, Nobuji Tetsutani
• Keywords: CHI 2004, Human Computer Interaction, Keypad Input, Text Entry Method, Unambiguous Input