Acoustic%20Source%20Direction%20by%20Hemisphere%20Sampling

1
Acoustic Source Direction by Hemisphere Sampling

• Stanley T. Birchfield
• Daniel K. Gillmor
• Quindi Corporation
• Palo Alto, California

2
The Problem
sound source
q
f
compact microphone array
3
Previous Work

• two microphones ? fHuang et al., ICASSP 1999
  Stephenne and Champagne, ICASSP 1995
• four microphones ? q, f Brandstein et al.,
  ICASSP 1995 Brandstein and Silverman, CSL 1997
• large microphone arrays ? r, q, fBrandstein et
  al., ICASSP 1995 Svaizer et al., ICASSP 1997

4
Our microphone array setup
microphone
d15cm
Note Arbitrary compact configurations can be
handled
5
Principle of Least Commitment
Principle of Least Commitment Delay decisions
as long as possible
Example
6
Previous Algorithm Cone Intersection
decision is made early
mic1 signal
prefilter
find peak
correlate
mic2 signal
prefilter
q,f
intersect
(may be no intersection)
mic3 signal
prefilter
correlate
find peak
mic4 signal
prefilter
Brandstein et al., ICASSP 1995 Brandstein and
Silverman, CSL 1997
7
Our Algorithm Sampled Hemisphere
map to common coordinate system
mic1 signal
prefilter
correlate
sampled hemisphere
mic2 signal
prefilter
correlate
final sampled hemisphere

correlate
q,f
combine
find peak
correlate
correlate
temporal smoothing
map to common coordinate system
mic3 signal
prefilter
correlate
mic4 signal
prefilter
decision is made after combining all the
available evidence
8
Pair-wise matching of signals
microphone
mic1 and mic2
mic3 and mic4
all four mics
9
Sampled Hemisphere
asymptotical cones (black regions are theoretical
blind spots, where cones have no
intersectionpractical blind spots are larger)
sampled hemisphere
10
Maximum error from matching non-coincident
microphones

• additional robustness outweighs possible error,
  which is negligible (less than 5 degrees
  when r gt 4d)

11
Experimental results(two trials, A and B)
A
A
A
B
B
B
Note also has been extensively used (hundreds
of hours) in real environments
12
Summary

• Advantages of sampled hemisphere algorithm
• handles arbitrary microphone array configurations
• has no blind spots
• demonstrates increased robustness, by
  followingthe principle of least commitment
• Future work
• integrate into a sound source location system
• investigate multiple simultaneous sound sources