We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.LG

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Sound

Title: End-to-End Sound Source Separation Conditioned On Instrument Labels

Abstract: Can we perform an end-to-end music source separation with a variable number of sources using a deep learning model? We present an extension of the Wave-U-Net model which allows end-to-end monaural source separation with a non-fixed number of sources. Furthermore, we propose multiplicative conditioning with instrument labels at the bottleneck of the Wave-U-Net and show its effect on the separation results. This approach leads to other types of conditioning such as audio-visual source separation and score-informed source separation.
Comments: 5 pages, 2 figures, 2 tables, ICASSP 2019
Subjects: Sound (cs.SD); Machine Learning (cs.LG); Audio and Speech Processing (eess.AS)
Cite as: arXiv:1811.01850 [cs.SD]
  (or arXiv:1811.01850v2 [cs.SD] for this version)

Submission history

From: Olga Slizovskaia [view email]
[v1] Mon, 5 Nov 2018 17:12:54 GMT (865kb,D)
[v2] Thu, 9 May 2019 16:55:38 GMT (943kb,D)

Link back to: arXiv, form interface, contact.