Bandwidth extension of musical audio signals with no side information using dilated convolutional neural networks

Abstract : Bandwidth extension has a long history in audio processing. While speech processing tools do not rely on side information, production-ready bandwidth extension tools of general audio signals rely on side information that has to be transmitted alongside the bitstream of the low frequency part, mostly because polyphonic music has a more complex and less predictable spectral structure than speech. This paper studies the benefit of considering a dilated fully convolutional neural network to perform the bandwidth extension of musical audio signals with no side information on the magnitude spectra. Experimental evaluation using two public datasets, medley-solos-db and gtzan, respectively of monophonic and polyphonic music demonstrate that the proposed architecture achieves state of the art performance.
Complete list of metadatas

Cited literature [21 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-02473457
Contributor : Mathieu Lagrange <>
Submitted on : Monday, February 10, 2020 - 5:07:56 PM
Last modification on : Thursday, February 13, 2020 - 1:35:19 AM

File

lagrangeBandwidthExtension.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02473457, version 1

Collections

Citation

Mathieu Lagrange, Félix Gontier. Bandwidth extension of musical audio signals with no side information using dilated convolutional neural networks. IEEE ICASSP, May 2020, Barcelona, Spain. ⟨hal-02473457⟩

Share

Metrics

Record views

3

Files downloads

8