University of Tokyo, Japan
Our team is comprised of three graduate students pursuing Computer Science at University of Tokyo. http://neurovoice.jp/
NeuroVoice is a voice converting software which converts your voice to anyone elses. NeuroVoice has three features. 1. NeuroVoice has the ability to convert individual’s voices to that of anyone else’s, even words that the target has never spoken. 2. NeuroVoice is a “voice to voice” conversion, not “text to voice”. Making it is possible to reflect the voice speed and its intonation onto the converted voice without losing the characteristics of the target voice. 3. Conversion is made possible by preparing audio data of just tens of minutes of the target person. NeuroVoice uses cutting edge Deep Learning method. When NeuroVoice learns data in the ten minutes, conversion is made possible. NeuroVoice has the ability to be applied to numerous fields, including movie dubbing with various individual voices, entertainment function of communication apps, mimicking songs, development of personalized speaking aids for speech-impaired subjects etc.