Paper Info
Title
HRTF aided broadband doa estimation using two microphones
Abstract
Two sensor broadband direction of arrival (DOA) estimation suffers from an inherent lack of dimensionality due to having just two sensors, yet humans and other animals are able to overcome this limitation using subtle variations introduced by the ears. Application of existing DOA estimation techniques to such systems becomes complicated due to the ill-behaved nature of the Head Related Transfer Function (HRTF). In this paper we present a subband signal extraction and focussing technique which retains the diversity information of the HRTF. We then develop a framework for combining these signals for subspace DOA estimation and investigate the constraints imposed on the single and multi-source DOA estimation problems. Finally, estimation performance is compared with existing techniques and we find performance has improved to be comparable to human localisation abilities.
Year
DOI
Venue
2012
10.1109/ISCIT.2012.6380863
ISCIT
Keywords
Field
DocType
audio signal processing,direction-of-arrival estimation,microphones,transfer functions,hrtf aided broadband doa estimation,direction of arrival estimation,diversity information,estimation performance,head related transfer function,human localisation,source localisation,subband signal extraction,subspace doa estimation,direction of arrival (doa),music,head related transfer function (hrtf)
Head-related transfer function,Computer science,Direction of arrival,Broadband,Real-time computing,Artificial intelligence,Audio signal processing,Computer vision,Subspace topology,Speech recognition,Curse of dimensionality,Transfer function,Signal extraction
Conference
ISBN
Citations 
PageRank 
978-1-4673-1155-7
0
0.34
References 
Authors
5
2
Name
Order
Citations
PageRank
Dumidu S. Talagala111.04
Abhayapala, T.D.217425.09