Non-linear ICA Based on Cramer-Wold Metric
Lists
Only you see this
Abstract
Non-linear source separation is a challenging open problem with many applications. We extend a recently proposed Adversarial Non-linear ICA (ANICA) model and introduce Cramer-Wold ICA (CW-ICA). In contrast to ANICA, we use a simple, closed–form optimization target instead of a discriminator–based independence measure. Our results show that CW-ICA achieves comparable results to ANICA while foregoing the need for adversarial training.
BibTeX
@inproceedings{spurek2020non,
author = {Przemyslaw Spurek and Aleksandra Nowak and Jacek Tabor and Lukasz Maziarka and Stanislaw Jastrzebski},
title = {{Non-linear ICA Based on Cramer-Wold Metric}},
booktitle = {ICONIP},
year = {2020},
url = {https://openreview.net/forum?id=dcKa8n4RJl},
}