Abstract :
A smart e-Receptionist is a computer application that is capable of sensing approaching visitors, orally greeting them, picking up their voice and understanding their questions, and then talking back to them appropriately. It may also have a limited self-learning capability of accumulating “knowledge” through its interactions with visitors. In this article, we present the recently developed smart e-Receptionist that integrates video-based motion detection (for sensing a nearby visitor), natural speech recognition (for understanding a visitor´s language), a question-answer (Q-A) knowledge database (for storing and managing information), and text-to-speech (TTS) (for a machine talking back to visitors). This e-Receptionist system can be easily converted into various information kiosks with slight modifications and configurations.
Keywords :
computer vision; human computer interaction; information management; information storage; interactive systems; natural language interfaces; natural language processing; object detection; question answering (information retrieval); speech recognition; speech synthesis; unsupervised learning; video signal processing; TTS; approaching visitor sensing; computer application; information kiosk; information management; information storage; knowledge accumulation; natural speech recognition; nearby visitor sensing; oral greeting; question understanding; question-answer knowledge database; self-learning capability; smart e-receptionist; talk back; text-to-speech; video-based motion detection; visitor interaction; visitor language understanding; voice pick up; Cameras; Databases; Keyboards; Motion detection; Speech recognition; Streaming media;
