Stone's Blind Source Separation Algorithm

Simple (and I mean simple - it's just one file with a single class) implementation of Stone's BSS algorithm. Heavily inspired by Dr. Stone's MATLAB Implementation.

Usage

Instantiate the StoneBSS class and pass the relevant hyperparameters. Then, call the .fit() method with the mixed signals as an argument. The StoneBSS instance can then be called like a function to unmix a given signal. Example usage can be found in stone_bss.py

References

@article{article,
author = {Stone, James},
year = {2001},
month = {08},
pages = {1559-74},
title = {Blind Source Separation Using Temporal Predictability},
volume = {13},
journal = {Neural computation},
doi = {10.1162/089976601750265009}
}