The SpeechBrain project aims to build a novel speech toolkit fully based on PyTorch. With SpeechBrain users can easily create speech processing systems, ranging from speech recognition (both HMM/DNN and end-to-end), speaker recognition, speech enhancement, speech separation, multi-microphone speech processing, and many others.

Anonymized SpeechBrain website for paper review

This project implements a **real-time speaker identification system** using the **ECAPA-TDNN** model from [SpeechBrain](https://speechbrain.github.io/). It enables users to speak into a microphone through a web interface, processes the audio in real time, and identifies the speaker based on pre-enrolled voice embeddings.