Transform-Invariant Non-Negative Matrix Factorization
Transform-Invariant Non-Negative Matrix Factorization
A comprehensive Python package for Non-Negative Matrix Factorization (NMF) with a focus on learning transform-invariant representations.
The packages supports multiple optimization backends and can be easily extended to handle application-specific types of transforms.
A general introduction to Non-Negative Matrix Factorization and the purpose of this package can be found on the corresponding GitHub Pages.
For using this package, you will need Python version 3.7 (or higher).
The package is available via PyPI.
Installation is easiest using pip:
pip install tnmf
The package comes with a streamlit demo and a number of examples that demonstrate the capabilities of the TNMF model.
They provide a good starting point for your own experiments.
Online Demo
Without requiring any installation, the demo is accessible via