VoiceCheck allows a chatbot to identify a user's voice to "remember" previous conversations!
Utilizing the principles of Mel-Frequency Cepstral Coefficients and Gaussian Mixture Models, VoiceCheck operates by analyzing a user's voice features & comparing it with stored voice biometrics to identify the speaker.
VoiceCheck identifies a user and continues previous conversations!
The inspiration for VoiceCheck came from observing the flaws of common voice assistants...These types of devices are generally tied to a singular account/profile, and operate with voice recognition based on what you say, and not how you say it...
VoiceCheck analyzes how you speak to determine your identity & proceeds to continue a conversation with you.
Utilizing the principles of Mel-Frequency Cepstral Coefficients and Gaussian Mixture Models, VoiceCheck operates by analyzing a user's voice features & comparing it with stored voice biometrics to identify the speaker.
VoiceCheck identifies a user and continues previous conversations!
The inspiration for VoiceCheck came from observing the flaws of common voice assistants...These types of devices are generally tied to a singular account/profile, and operate with voice recognition based on what you say, and not how you say it...
VoiceCheck analyzes how you speak to determine your identity & proceeds to continue a conversation with you.
Imagine Amazon's Alexa continuing actual conversations with you!
VoiceCheck Prototype
SIP Presentation