The Sonic Eye: A device for ultrasonic human echolocation

The Sonic Eye: A device for ultrasonic human echolocation
The Sonic Eye: A device for ultrasonic human echolocation
Santani Teng1, Jascha Sohl-Dickstein2, Alex Maki-Jokela, Benji Gaub1, Mike DeWeese1, Chris Rodgers1, & Nicol Harper3
1University of California, Berkeley 2Stanford University 3University of Oxford
Contact: [email protected]; www.thesoniceye.com
Introduction: Echolocation
Many non-human animals, and even some blind
humans, use active echolocation vocalizations to
aid in navigation and object perception.
The Sonic Eye presents users with rich ultrasonic
echolocation cues that are not otherwise available
to humans, or difficult to control.
Sonic Eye Prototype
Emitter/Mic Power Cones
Single centrally placed
ultrasonic emitter
Stereo ultrasonic
microphones (flat
frequency response to
100 kHz)
Artificial pinnae
(mimicking bat ears)
capture elevation
information
PC generates
waveforms,
manages GUI and
processes echo
input
Minimal processing of cues exploits the auditory
system’s inherent ability to process spatial and
object information.
Time-stretched
output through
stereo earbuds
Information Flow Overview
Preliminary Behavioral Data
In development:
Controller allows
customization of
pulse type, emission
rate, etc.
One sighted blindfolded volunteer localized a 30cm
disk held at various positions at a 1m distance.
Pulse & Echo Spectrograms
Ear bud output (time-dilated chirp + echo)
(Reflecting
object, 150cm
distance)
x 10
Noise
33 44
5 66 77
5
Time(s)
88 x9910
7
7
6
6
5
4
3
2
4
1
11
x 10
22
33 44
5 66 77
5
Time(s)
88 x9910
4
9
9
8
8
7
7
6
6
5
4
0.02
20
0.03
0.04
30
Time(s) 40
0.05
50
0.06
60
22
33 44
5 66 77
5
Time(s)
9
88 9
x 10
Future Directions
4
3
2
2
1
0.01
10
0.02
20
0.03
0.04
30
Time(s) 40
0.05
50
Time (ms)
0.06
60
0
−3
4
5
3
0
11
x 10
1
0.01
10
0
−3
Responses
Detection was perfect (absences always reported).
Localization was above chance, with 72% of
responses within one sector of the correct location.
4
2
0
Configuration (100 trials)
5
3
−3
Chirp
8
Frequency(Hz)
22
4
9
Frequency(Hz)
Frequency(Hz)
8
1
11
x 10
909
808
707
606
505
404
303
202
101
00
4
9
Frequency(Hz)
E
C
H
O
x 10
Click
Frequency(Hz)
Microphone input (chirp & echoes)
P
U
L
S
E
4
Frequency(Hz)
Speaker output (e.g. chirp, click)
Frequency (kHz)
x 10
909
808
707
606
505
404
303
202
101
00
0.01
10
0.02
20
0.03
0.04
30
Time(s) 40
0.05
50
0.06
60
We will characterize the envelopes of object detection,
localization, discrimination, and navigation; refine the
user interface for on-line user customization of echo
parameters (e.g. tone, rate, intensity); and move to more
robust, compact hardware and software platforms.
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement