n this presentation, Didier will present demo based voice recognition and pitch detection. He will introduce the new VoiceGesture API and explain how to create a sound model, a voice group, and how to catch voice events.
Didier will also discuss how to build a SingStar-like application from scratch, prepare a mp3-pitch-track, and detect the pitch in real-time. Finally, he will take us through an in-depth explanation of the Dynamic Time Warping algorithm, explaining how we can compare two given images representing data that vary in time.
As a gift, all participants of this session will receive a cd-rom containing the first public release of VoiceGesture and all the sources used during the session.