Distributed FastMNMF imposes block-diagonal structure on source spatial covariance matrices to perform local inversions and share NMF spectrogram models across subarrays for faster, scalable blind source separation.
Exploiting speaker embeddings for improved microphone clustering and speech separation in ad-hoc microphone arrays,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
eess.AS 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Fast Multichannel NMF with Block-Diagonal Spatial Covariance Matrices for Efficient Blind Source Separation Using Distributed Microphone Arrays
Distributed FastMNMF imposes block-diagonal structure on source spatial covariance matrices to perform local inversions and share NMF spectrogram models across subarrays for faster, scalable blind source separation.