This repository contains code for the papers
Kristina Tesch, Nils-Hendrik Mohrmann, and Timo Gerkmann, "On the Role of Spatial, Spectral, and Temporal Processing for DNN-based Non-linear Multi-channel Speech Enhancement", accepted at Interspeech 2022, [arxiv], [audio examples]
and
Kristina Tesch and Timo Gerkmann, "Insights into Deep Non-linear Filters for Improved Multi-channel Speech Enhancement", submitted to Transactions of Audio, Speech and Language Processing, [audio examples]