Multi-speaker separation employing microphone array and vertex finding algorithm

© 2018 IEEE. This paper proposes a new speaker detection and signal separation algorithm for multiple speakers using microphone array data recorded in a room environment. The algorithm utilizes the fact that in multi-speaker conversations not all speakers are speaking simultaneously there are time s...

Full description

Bibliographic Details
Main Authors: Hong Dam, H., Nordholm, Sven
Format: Conference Paper
Published: 2018
Online Access:http://hdl.handle.net/20.500.11937/74645
_version_ 1848763333090476032
author Hong Dam, H.
Nordholm, Sven
author_facet Hong Dam, H.
Nordholm, Sven
author_sort Hong Dam, H.
building Curtin Institutional Repository
collection Online Access
description © 2018 IEEE. This paper proposes a new speaker detection and signal separation algorithm for multiple speakers using microphone array data recorded in a room environment. The algorithm utilizes the fact that in multi-speaker conversations not all speakers are speaking simultaneously there are time segments when only a single speaker is active. Based on that observation a speech activity detector for each speaker (MVAD) has been developed. It is based on SRP-PHAT estimates for different blocks of data. We have shown that these estimates form vertexes in a convex polygon which can be employed to obtain MVAD detections. Those detections are then used to form Minimum Variance Distortionless Response (MVDR) beamformers. Evaluations based on real recorded speech data with 4 speakers show that the algorithm provides good interference suppression and low speech distortion for this difficult scenario.
first_indexed 2025-11-14T11:01:47Z
format Conference Paper
id curtin-20.500.11937-74645
institution Curtin University Malaysia
institution_category Local University
last_indexed 2025-11-14T11:01:47Z
publishDate 2018
recordtype eprints
repository_type Digital Repository
spelling curtin-20.500.11937-746452019-02-19T05:35:45Z Multi-speaker separation employing microphone array and vertex finding algorithm Hong Dam, H. Nordholm, Sven © 2018 IEEE. This paper proposes a new speaker detection and signal separation algorithm for multiple speakers using microphone array data recorded in a room environment. The algorithm utilizes the fact that in multi-speaker conversations not all speakers are speaking simultaneously there are time segments when only a single speaker is active. Based on that observation a speech activity detector for each speaker (MVAD) has been developed. It is based on SRP-PHAT estimates for different blocks of data. We have shown that these estimates form vertexes in a convex polygon which can be employed to obtain MVAD detections. Those detections are then used to form Minimum Variance Distortionless Response (MVDR) beamformers. Evaluations based on real recorded speech data with 4 speakers show that the algorithm provides good interference suppression and low speech distortion for this difficult scenario. 2018 Conference Paper http://hdl.handle.net/20.500.11937/74645 10.1109/IWAENC.2018.8521373 restricted
spellingShingle Hong Dam, H.
Nordholm, Sven
Multi-speaker separation employing microphone array and vertex finding algorithm
title Multi-speaker separation employing microphone array and vertex finding algorithm
title_full Multi-speaker separation employing microphone array and vertex finding algorithm
title_fullStr Multi-speaker separation employing microphone array and vertex finding algorithm
title_full_unstemmed Multi-speaker separation employing microphone array and vertex finding algorithm
title_short Multi-speaker separation employing microphone array and vertex finding algorithm
title_sort multi-speaker separation employing microphone array and vertex finding algorithm
url http://hdl.handle.net/20.500.11937/74645