An end-to-end Speaker Recognition System written in Python using Mel Frequency Cepstral Coefficients (MFCC) and Spectral Subband Spectroid (SSC) for features. A Mini Project requirement for CS 280 Intelligent Systems course at University of the Philippines Diliman AY 2015-2016 under Sir Prospero Naval.

• Binary SVM per user
• One vs Rest Multi Class SVM
• Decision Tree
• Gaussian Naive Bayes Classifier

• Python 2.7.10
• Microphone/Audio Input System capable of recording CD Quality (Stereo channels, 16 bit, 44100 Hz)

The following Python libraries not yet included in Python 2.7.10 Standard distribution needs to be installed:

• pyaudio
• tkintertable
• wave
• scipy
• libsvm
• scikit-learn
• Python Speech Features by James Lyons

After installing required Python libraries

1. Clone and download the repository.
2. Go to the downloaded directory and run the application

python speaker-recognizer.py