Wide-Range Pitch Detection
Detecting The Pitch of Speech and Music
Over The Full Range of Human Hearing.
Modern AI systems focus heavily on words and text. But in speech, singing, and music, the pitch of a sound is a vital quality that carries meaningful information about emotion and beauty. Merely decoding the words leaves us without a meaningful understanding of what we are hearing.
References
Lloyd Watts (2018), "DeepPitch: Wide-Range Monophonic Pitch Estimation using Deep Convolutional Neural Networks."