1 comment

[ 3.2 ms ] story [ 12.4 ms ] thread
Paper: http://web.media.mit.edu/~mehoque/Publications/13.Hoque-etal...

I found the choices of software particularly interesting:

"Prosody Analysis: ... To extract prosodic features, we developed an application programming interface (API) using the low level signal processing algorithms included in the Praat [praat.org], an open source speech processing toolkit."

"Speech Recognition: For real time speech recognition, we used the Nuance speech recognition software development kit (SDK). ..."