Implementation of "SpEx: Multi-Scale Time Domain Speaker Extraction Network".

This is a demo for my bachelor thesis 'Speaker Separation and Machine Auditory Perception for Dialogue Scene'.

Stream Server for connecting Twilio's Media Stream to Symbl over a WebSocket with an exposed RESTful API for triggering the delivery of Symbl’s real-time events to a Client server.