Top

WASPAA papers on source localisation & source separation

I am delighted that three of our papers were accepted for WASPAA 2021!

Aidan Hogg’s work [1] proposes to combine spectral features (voice pitch) and spatial features (source direction-of-arrival) in order to track multiple, competing speakers in complex scenarios.

Vincent Neo’s work [2] proposes to exploit spatial, temporal – and hence spectral – diversity of competing sound sources to decorrelate multiple, overlapping sound events.

Hsuan-Yang Wang’s work [3] proposes a new set of physiologically inspired acoustic features that mimic the human auditory process for source localisation using deep neural networks.

Looking forward to seeing you at WASPAA in Oct’2021.


[1] Hogg, Aidan, Neo, Vincent W., Weiss, Stephan, Evers, Christine and Naylor, Patrick (2021) A polynomial eigenvalue decomposition MUSIC approach for broadband sound source localization. In IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA). New Paltz, United States. 17 – 20 Oct 2021. 5 pp . (In Press)

[2] Neo, Vincent W., Evers, Christine and Naylor, Patrick (2021) Polynomial matrix eigenvalue decomposition-based source separation using informed spherical microphone arrays. In IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA). New Paltz, United States. 17 – 20 Oct 2021. 5 pp . (In Press)

[3] Wang, Hsuan-Yang, Nelson, Philip and Evers, Christine (2021) Excitation-inhibition Cell activity patterns for binaural source localisationIn IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA). New Paltz, United States. 17 – 20 Oct 2021. 5 pp . (In Press)

No Comments

Post a Comment