with Upgraded LIUM and sphinx code
This app analyzes the conversation directly your device to identify who is speaking and when. It does this with no effort on your part by using modern machine learning principles applied to the sounds recorded from your microphone. The analysis report after a conversation provides the total time each person spoke. Also, a scrollable piano plot is provided to observe the dynamics between speakers throughout the conversation.
This app is a prototype effort by students working with Dr. Ting Xiao in the Loyola University Chicago Computer Science department. It primarily relies on the LIUM package to diarize the audio signal. Initial development was performed by Marcos Leal and Edgard Oliveira with later improvements by Albert Du and final development and deployment by Rejoice Jebamalaidass.
