ETRI-Knowledge Sharing Plaform

KOREAN
논문 검색
Type SCI
Year ~ Keyword

Detail

Conference Paper Sound Source Localization using Multiple Microphones Attached on the Wearable Glasses
Cited - time in scopus Share share facebook twitter linkedin kakaostory
Authors
Kang-Ho Park, Yi-Gyeong Kim, Sung-Q Lee, Bohyun Kim, Seoyoung Lee, Hyung-Min Park
Issue Date
2018-05
Citation
International Conference on NOise and Vibration Emerging Methods (NOVEM) 2018, pp.1-9
Language
English
Type
Conference Paper
Abstract
The sound source localization using multiple microphones attached on the wearable glasses is investigated for the recognition of the location of the alarming sounds or voices. The sound signals are measured by small size MEMS microphones attached on the glasses placed on a mannequin, and the sound source location such as a distance from the sound source as well as the azimuthal angle is obtained and analyzed by steered response power with the phase transform (SRP-PHAT) and the probabilistic Gaussian mixture model (GMM) algorithm based on the time difference of arrival between the microphone pairs. The estimation accuracies of the azimuthal angle and the distance from the sound source are evaluated and analysed with both GMM and SRP-PHAT methods. We found that GMM method is better than SRP-PHAT for sound source localization, and the distance as well as the azimuthal angle can be well estimated by both the simulation and experimental work. This technology can be applied to the safety assistance service for the handicapped persons in the recognition of the environmental alarming sounds.
KSP Keywords
Experimental work, GMM method, Gaussian Mixture Models(GMM), Gaussian mixture(GM), MEMS Microphone, Phase transform, SRP-PHAT, Source Location, Steered response power, Time Difference of Arrival(TDOA), azimuthal angle