Localization of multiple speakers based on a two step acoustic map analysis

Abstract

An interface for distant-talking control of home devices requires the possibility of identifying the positions of multiple users. Acoustic maps, based either on Global Coherence Field (GCF) or Oriented Global Coherence Field (OGCF), have already been exploited successfully to determine position and head orientation of a single speaker. This paper proposes a new method using acoustic maps to deal with the case of two simultaneous speakers. The method is based on a two step analysis of a coherence map: first the dominant speaker is localized; then the map is modified by compensating for the effects due to the first speaker and the position of the second speaker is detected. Simulations were carried out to show how an appropriate analysis of OGCF and GCF maps allows one to localize both speakers. Experiments proved the effectiveness of the proposed solution in a linear microphone array set up.Index Terms-microphone array, speaker localization, multiple speakers, global coherence field.

A Novel Nested Circular Microphone Array and Subband Processing-Based System for Counting and DOA Estimation of Multiple Simultaneous Speakers

“…For the second sub-array [i.e., microphone pairs (1,3), (3,5), (5,7), (7,1), (2,4), (4,6), (6,8), (8,2)], the distance between microphones is set to 5.5 cm. For the third sub-array [i.e., microphone pairs (1,4), (2,5), (3,6), (4,7), (5,8), (6,1), (7,2), (8,3)], we adjust the inter-microphone distance to 7.8 cm. Finally, for the farthest pairs of microphones [i.e., microphone pairs (1,5), (2,6), (3,7), (4,8)], the distance between microphones is set to 10.5 cm.…”

“…For the third sub-array [i.e., microphone pairs (1,4), (2,5), (3,6), (4,7), (5,8), (6,1), (7,2), (8,3)], we adjust the inter-microphone distance to 7.8 cm. Finally, for the farthest pairs of microphones [i.e., microphone pairs (1,5), (2,6), (3,7), (4,8)], the distance between microphones is set to 10.5 cm. In this way, spatial aliasing does not happen in microphone signals in any of sub-arrays.…”

“…Then the GCC function is calculated for corresponding pairs of microphones in each subbands. For example, for the 4-5, 5-6, 6-7, and 7-8 kHz bands, the GCC function is calculated only for microphone pairs (1,2), (2,3), (3,4), (4,5), (5,6), (6,7), (7,8), (8,1) and the other microphone pairs are not used in this band. In this way, the spatial aliasing is considerably eliminated.…”

“…methods based on interaural time difference (ITD). The second category contains the methods based on energy and spatial search for speaker localization [3,15,23]. In this category, the energy at candidate points of space is checked to determine the points which are most likely to be a sound source.…”

On-Board Relative Bearing Estimation for Teams of Drones Using Sound

“…The Coherence pruning unit is used to also obtain the bearing information of other existing robots that are potentially masked behind the sound of the dominant robot. Inspired by the work of [36], the idea is to This unit initially uses Equation (7) to compute the time delay τ bmij , for every microphone pair, corresponding to the obtained bearing measurement − → b m . This time delay is then used to generate a pruning sequence U ij (τ ) that is added to the similarity measurements R ij (τ ) to de-emphasize the existing peak of the dominant source at time lag τ bmij .…”