VoMIS - the VoiceXML-Based Multimodal Interactive System for NAO Robot

The proposed paper brings a description of the VoiceXML-based multimodal interactive system (VoMIS) for NAO humanoid robot. The designed system enables the multimodal interaction with the user in such manner that it takes a speech input from the user and it answers by a combination of synthetic speech and gestures. The core of the system is an external dialogue manager VoiceON, which interprets VoiceXML language. VoiceXML was originally designed for unimodal systems, but thanks to its advantages we decided to extent it to manage multimodal interactions. Our work illustrates how VoiceXML can be easily extended to manage also multimodal interaction, mainly using <prompt> element. Designed changes enables to control movements and gestures of the robot.