Deep-Learning-Speaker Classification

Project Description

Project to learn about speech recognition - both Speaker Diarization and other Speech Recognition applications.

The first objective will be to implement Speaker Classification using an SVM.

Current Data Set

The data being used for this project can be found at: http://www.openslr.org/12/ . Openslr is an open source project dedicated to hosting speech and language resources, hoping to assist the progress of speech recognition.

Previous Work Done

1. Pannous is a project that is working on implementing Speech Recognition in Google's Tensorflow.

   Github link: https://github.com/pannous/tensorflow-speech-recognition/

   To see more of an in-depth walkthrough of how Pannous approaches the speaker classification problem, please see Pannous-Walkthrough.md.

Installation Requirements

- Librosa

- Pydub

- TFLearn

Github Navigation