Open Speech Platform User Guide OSP Manual
User Manual:
Open the PDF directly: View PDF .Page Count: 34
P R O J E C T S P O N S O R N AT I O N I N S T I T U T E O F H E A LT H , N I D C D .
T H I S W O R K I S S U P P O R T E D B Y T H E N AT I O N A L I N S T I T U T E O N D E A F N E S S
A N D O T H E R C O M M U N I C AT I O N D I S O R D E R S , O F T H E N AT I O N A L I N S T I T U T E O F
H E A LT H ( N I D C D / N I H ) . R 0 1 D C 0 1 5 4 3 6 : " A R E A L - T I M E , O P E N , P O R TA B L E , E X TENSIBLE SPEECH LAB" TO THE UNIVERSITY OF CALIFORNIA, SAN DIEGO.
H A R I N AT H G A R U D A D R I
O P E N S P E E C H P L AT F O R M
USER GUIDE
OSP TEAM
Copyright © 2019 Harinath Garudadri
published by osp team
openspeechplatform.ucsd.edu
Copyright 2017 Regents of the University of California
Redistribution and use in source and binary forms, with or without modification, are permitted provided
that the following conditions are met:
1. Redistributions of source code must retain the above copyright notice, this list of conditions and the
following disclaimer.
2. Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the
following disclaimer in the documentation and/or other materials provided with the distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT,
INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA,
OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF
THE POSSIBILITY OF SUCH DAMAGE.
First printing, June 2019
Contents
Abstract
8
Setting up the RT-MHA
2.1 Installing the Software
9
9
Release 2019a Test and Validation
2.2 Connecting Audio I/O Devices
2.2.1
On OS X machines
11
2.2.2
On Linux machines
12
2.3 Setting up Audio I/O
11
11
13
2.3.1
Internal Audio Codecs
2.3.2
External USB Audio Codecs
13
14
Verifying the RT-MHA is Working Properly
3.1 Internal Audio Codecs
16
16
3.1.1
Testing with audio files
3.1.2
Live Testing with headphones
3.1.3
Live Testing with Andrea SB-405 Super beam headphones
16
17
RT-MHA Command line arguments
4.1 OSP commands
4.1.1
-h for Help
18
18
18
17
5
4.1.2
-F for choosing Sampling frequency
4.1.3
-r to enable/ disable Rear mics
4.1.4
-a for enabling/disabling Adaptive Feedback cancellation (AFC)
4.1.5
-d to set the attenuation factor
4.1.6
-T for selecting AFC algorithm
4.1.7
-N for choosing noise estimation type
4.1.8
-S for enabling/disabling spectral subtraction
4.2 OSP modes of control
18
19
19
20
TCP daemon interface
4.2.2
Playback mode interface
4.2.3
Playback mode with TCP daemon interface
4.2.4
Command Line interface
20
21
21
21
4.3 Recommended Settings for ANSI 3.22 Testing
4.4 Recommended Settings for Testing with Webserver
Setting up the Embedded Web Server (EWS)
5.1 Running RT-MHA on Local Computer
5.2 Connecting to OSP Remotely
Verifying the EWS is Working
6.2 4AFC
Bibliography
31
32
33
20
20
4.2.1
6.1 Researcher Page
19
24
31
24
22
22
23
19
List of Figures
2.1 Terminal output after the ./install command. You will need to enter your password at various times. snip represents a large portion
of the terminal output deleted in this figure. This process can take
30 − 90 minutes, depending on your computer and network speed.
The script ends with instructions on invoking osp and ews in the terminal.
10
USB Audio Input Configuration.
12
USB Audio Input Configuration.
12
List of devices displayed by running the pa_devs command
13
This is what your finder window will look like when you’re looking
up Audio MIDI Setup.
13
2.6 This is what the Audio MIDI Setup panel should look like when you
select this device for sound output.
14
2.7 This is what the System Preferences>Sound should look like when
you select this device for sound output.
14
2.8 This is what the System Preferences>Sound should look like when
you select this device for sound input.
15
2.2
2.3
2.4
2.5
5.9 "This is what your Finder window should look like when you have
selected your Hard Drive
25
5.10 Here are all the files you have made visible with this command, with
the etc folder highlighted
25
5.11 Here are all the files you have made visible with this command, with
the etc folder highlighted
26
5.12 This is how you Open the httpd.conf file using Atom
26
5.13 This is what it should look like after you delete the # on the appropriate line
27
5.14 This is what it should look like after you delete the # on the appropriate line
27
5.15 This is what your Terminal window will look like when you use the
whoami command. The output is highlighted. Your output will be
your username, not tzubatiy.
28
7
5.16 This is what your Atom window should look like when have added
your name to the user line, in this case this is my name.
28
5.17 This is what your Atom window should look like when have added
these lines of code in the right place.
28
Abstract
This version corresponds to Release 2019a. We present all of the
necessary tools that need to be installed to make the Open Speech
Platform (OSP) run smoothly, and how to make sure they are working properly. The OSP is comprised of a real time master hearing
aid (RT-MHA) portion and an Embedded Web Server (EWS) portion. This guide is organized based on these two chapters, so first we
present the tools to get the RT-MHA all set up, and then we present
the setup process for the EWS.
https://github.com/nihospr01/OpenSpeechPlatform-UCSD.
Setting up the RT-MHA
2.1
Installing the Software
1. Download 2018c. Download the latest release from https://
github.com/nihospr01/OpenSpeechPlatform-UCSD. You can either
“clone the software on your computer” or “download ZIP file to
your computer.”
(a) If you wish to clone, type cd in to your terminal, to make sure
you are in the local directory. Then press clone in github, and
copy the line of code generated into your terminal. Once it has
cloned, type cd nameofdirectory (for example cd OpenSpeechPlatformUCSD). Then, type git pull to get the latest version. This is the
preferred approach.
(b) When you download a .zip file from github, you will have to
manually download newer versions with bug fixes.
2. Install 2018c Open a terminal. cd ./Software/ and ./install.
This script does several things.
(a) Identify the operating system (OS) on your computer – currently OS X and Linux.
(b) Install software packages including portaudio (for realtime
audio input/output); MySQL (a relational database), PHP (a
server side scripting language) and other packages for LAMP
software stack (LAMP stands for Linux, Apache, MySQL and
PHP).
(c) Build RT-MHA and EWS.
(d) Finally, it installs osp in /usr/local/bin/osp and a script to
invoke ews in /usr/local/bin/ews.
If everything went well, your screen will look similar to Figure 2.1.
10
Figure 2.1: Terminal output after the
./install command. You will need to
enter your password at various times.
snip represents a large portion of the
terminal output deleted in this figure.
This process can take 30 − 90 minutes,
depending on your computer and
network speed. The script ends with
instructions on invoking osp and ews in
the terminal.
install.png
Release 2019a Test and Validation
This chapter describes sanity tests to validate your versions of osp
and ews.
2.2
Connecting Audio I/O Devices
There are many audio input/output options for OSX and Linux
computers. One inexpensive option is the Andrea Communications
3D Surround Sound Recording CANS. They are supra-aural headsets,
with left and right mics.
You can also use a high end audio device such as Zoom TAC-8 or
Focusrite Scarlett 2i2.
2.2.1
On OS X machines
1. Plug the headsets in to an available USB port.
2. On Mac computers, Open Audio MIDI Setup: This can be found
in Finder | Applications | Utilities, as shown in Figures 2.2
and 2.3.
3. Select 48,000 Hz option on both screens and levels to 1..0 as
shown.
4. If you use a high end audio interface box (such as Zoom TAC-8)
and if you have 98,000 Hz and 24 bit option, you should choose
this.
12
Figure 2.2: USB Audio Input Configuration.
Figure 2.3: USB Audio Input Configuration.
2.2.2
On Linux machines
1. On Linux machines the default audio is not set. Run the command
pa_devs to give the list of the available devices and the device
numbers they correspond to
2. Run the command osp -input_device x -output_device y,
where x and y are values attained from the previous command.
With reference to Figure 2.4, the command issued is osp -input_device
6 -output_device 6
13
Figure 2.4: List of devices displayed by
running the pa_devs command
Can we use white background in the terminal shots for the
document? Black is good for code development, but I am not sure for
documents
2.3
2.3.1
Setting up Audio I/O
Internal Audio Codecs
It is possible to test proper functioning of RT-MHA using the internal
audio codecs built into a MacBook.
1. Open Audio MIDI Setup: This can be found in Finder: Applications > Utilities, as shown in Figure 2.5.
Figure 2.5: This is what your finder
window will look like when you’re
looking up Audio MIDI Setup.
2. Select primary input: In Audio MIDI Setup, right-click (CTRL+click)
Built-in Microphone on the left hand side. Select "Use This Device
14
For Sound Input. Change the format to 96,000 Hz."
2.3.2
External USB Audio Codecs
1. Select primary output: In Audio MIDI Setup, click on the Zoom
Tac-8, click on Input, select the format 96,000 Hz. Then click
Output, and also select 96,000 Hz."
Figure 2.6: This is what the Audio MIDI
Setup panel should look like when you
select this device for sound output.
2. Edit Sound Preferences: Open System Preferences>Sound and
select the Output tab. Click on Zoom Tac 8. Select the Input tab
and also click on Zoom Tac 8.
Figure 2.7: This is what the System
Preferences>Sound should look like
when you select this device for sound
output.
15
Figure 2.8: This is what the System
Preferences>Sound should look like
when you select this device for sound
input.
Verifying the RT-MHA is Working Properly
3.1
3.1.1
Internal Audio Codecs
Testing with audio files
In this section, we will use audio samples provided with the OSP
software to test the Real time Master Hearing Aid.
1. In OSP we have provided 96kHz sample speech files for testing
the software. All of this is found in the following directory: OSP >
Matlab > InputFiles-96kHz.
2. On the terminal run the following command to make OSP process
audio files and play them through the speaker. Here we use the -p
argument to play sound through the software.
osp -F1 -M1 -a -r -d-40 -p ~/Downloads/OSP/matlab/Input\ Files-32k/TIMIT_female_sa2.wav
This would play the audio file "TIMIT_female_sa2.wav" through
the entire OSP process, and you would hear the sound through the
speaker/headphones.
3. We have also provided the option of playing the sounds with NH,
N2, N4 and S2 settings. These can be executed using the following
command:
osp -F1 -M1 -a -r -d-40 -p ~/Downloads/OSP/matlab/Input\ Files-32k/TIMIT_female_sa2.wav -s1
Here the argument -s gives control to the user
to choose the gain settings for OSP and play the audio with those
settings. Setting_number are set the following way: 1= N2, 2 = N4,
3 = S2 and 0 = NH.
17
3.1.2
Live Testing with headphones
1. For live testing you need a headset with microphone. We found
headphones commonly used for making calls to be adequate. Plug
this into your computer.
2. Make sure you have set the volume on your computer to the
maximum allowed.
3. Enter the following command on the terminal to start osp in
command line interface mode.
osp -F1 -M1 -a -r
Here choose F1 if your system supports 96kHz or else choose F0
for 48kHz.
4. Next you can increase the gain to 10dB on all bands by typing in
"g10". This way your own voice would be heard louder through
the headphones.
3.1.3
Live Testing with Andrea SB-405 Super beam headphones
1. For live testing you need a headset with microphone. We found
headphones commonly used for making calls to be adequate. Plug
this into your mac at the USB port.
2. Make sure the mute switch is in the "off" position.
3. Make sure you have set the volume on your computer to the
maximum allowed.
4. Enter the following command on the terminal to start osp in
command line interface mode.
osp -F1 -M1 -a -r
Here choose F1 if your system supports 96kHz or else choose F0
for 48kHz.
5. Next you can increase the gain to 10dB on all bands by typing in
"g10". This way your own voice would be heard louder through
the headphones.
RT-MHA Command line arguments
In this section we explore the various functions of the RT-MHA on
the command line interface. There are multiple starting configurations and these configurations are specified using arguments.
4.1
4.1.1
OSP commands
-h for Help
This would show all the available options/configurations the RTMHA can be executed in, as shown Figure??.
osp -h
4.1.2
-F for choosing Sampling frequency
The first thing you should do is choose your sampling frequency
using the extension -F This should correspond with the sampling
frequency you picked in step 1.11 and 1.12. The Sampling rates
supported are 0, 48kHz and 1, 96kHz.
To set the sampling rate to 48kHz, type the command below into
your terminal and press Enter:
osp -F0
Notice how in the terminal window it now says:"Choosing 48kHz
sampling frequency"
To stop running OSP, type on the keyboard "control"+"c" do
this now, and we will continue verifying that everything is working properly. You will need to do this step after every step in this
section
To set the sampling rate to 96kHz, type the command below into
your terminal and press Enter:
osp -F1
19
Notice how in the terminal window it now says:"Choosing 96kHz
sampling frequency"
Type on the keyboard "control"+"c" to stop running OSP before
going on to the next step.
4.1.3
-r to enable/ disable Rear mics
Open a Terminal window, type in the line below and press Enter
osp -r
This is the recommended default setting as we do not have rear
mic support yet. Notice how in the terminal window, it says "Disabling rear mics".
Type on the keyboard "control"+"c" to stop running OSP before
going on to the next step.
4.1.4
-a for enabling/disabling Adaptive Feedback cancellation (AFC)
In a Terminal window, type in the line below and press Enter:
osp -r -a
You are now running OSP from the command line with the rear mics
disabled and AFC off. On the first line in the terminal window it now
says: "4AFC has been disabled" Type on the keyboard "control"+"c"
to stop running OSP before going on to the next step.
4.1.5
-d to set the attenuation factor
To run OSP with an attenuation factor (let’s say -40dB attenuation),
type the line below into a terminal and press Enter:
osp -r -a -d -40
you can also not add a space between the -d and the -40 as such:
osp -r -a -d-40
Notice how in the terminal window it now says: " Attenuation factor is -40.000000 dB" Type on the keyboard "control"+"c" to stop
running OSP before going on to the next step.
4.1.6
-T for selecting AFC algorithm
To start OSP with a predefined AFC type use the extension -T
The types of AFC supported are ’0’ = FXLMS, ’1’ = PNLMS and ’2’
= SLMS. For example, say you want to set the setting to SLMS. Then
you would type the line below into your terminal and press Enter:
20
osp -T2 -r
Notice how in the terminal window it now says: Setting adaptation
type to SLMS Type on the keyboard "control"+"c" to stop running
OSP before going on to the next step.
4.1.7
-N for choosing noise estimation type
To run OSP with a predefined Noise Estimation algorithm, use the
extension -N-noiseExtension
The types of Noise estimation algorithms RT-MHA supports are
’1’ = Arslan Power Averaging Procedure, ’2’ = Hirsch and Ehrlicher
weight averaging procedure and ’3’ = Cohen and Berdugo MCRA
Procedure.
For example, say you wanted to set the Noise Estimation algorithm to the Cohen and Berugo. To do this, type the line below into
your temrminal and press Enter:
osp -N3 -a -r
Notice how in the terminal window it now says: Cohen and Berdugo
MCRA Procedure Noise estimation method chosen Type on the
keyboard "control"+"c" to stop running OSP before going on to the
next step.
4.1.8
-S for enabling/disabling spectral subtraction
To run OSP with spectral subtraction, use the extension -S1 or S0 By setting spectral subtraction to ’-S0’, spectral subtraction is
disabled; by setting to ’-S1’, spectral subtraction is enabled.
osp -S1 -a -r
Notice how in the terminal window it now says: Spectral subtraction
enabled. Type on the keyboard "control"+"c" to stop running OSP
before going on to the next step.
4.2
4.2.1
OSP modes of control
TCP daemon interface
-t starts the application in TCP daemon mode. In this mode, the
application is waiting for a connection on port 8001 from an OSP
client. Once a client is connected, all parameters (such as -g50 and
-g80 gains, attack/release time, knee points, etc) come from the client.
21
The command line version of the application cannot be changed
manually. To stop the application once it’s started in TCP mode, the
user must hit CTRL+C.
osp -t
Notice how in the terminal window it now says:"Using TCP daemon
mode" and "waiting for a connection from client..."
4.2.2
Playback mode interface
-p starts the application in playback-only mode. The
audio is not enabled, and instead, audio data is read from into the application where the RT-MHA algorithm is
applied to the audio data, and this is played out in the hearing aid/
speakers/ headphones. This mode can be used for verifying the
functionality of the software.
osp -p ~/Downloads/OSP/matlab/Input\ Files-32k/TIMIT_female_sa2.wav
Notice how in the terminal window it now says:"Playback mode specified, setting up file /Users/kc/Downloads/OSP/matlab/InputFiles32k/TIMIT_female_sa2.wav"
4.2.3
Playback mode with TCP daemon interface
-l starts the application in TCP daemon and
4.2.4
Command Line interface
When neither of -t, -l or -p have not been specified.
OSP can be interacted with on the command line. The following
commands are can be used to interact with OSP on the command line
(terminal).
1. ’g ’: If the operator chooses to adjust the gain to
10dB, he/she can simply enter ’g 10’ followed by Enter on the
command line.
2. ’s’ If the operator wants to turn on/off AFC, he/she can simply
enter ’s’ followed by enter to toggle AFC mode.
3. ’p ’ If the operator chooses to
adjust the spectral subtraction to 0.5, he/she can simply enter ’p
0.5’ followed by Enter on the command line. The range for spectral
subtraction parameter is from 0 to 1.
4. ’c’ to save the converged afc filter coefficients.
22
5. ’f’ to toggle feedback.
6. ’q’ to quit the application.
4.3
Recommended Settings for ANSI 3.22
Testing
For ANSI 3.22 testType the following line of code into your Terminal
and press Enter:
osp -a -r -M0 -d-40
Here the attenuation factor is -40. Please use -40dB when using with
the breakout board (bob) and jellybeans.
4.4
Recommended Settings for Testing with
Webserver
textFor testing with webserver, Type the following line of code into
your Terminal and press Enter
osp -t -a -r -M1 -d-40
Setting up the Embedded Web Server
(EWS)
1. Enter the following line of code into your Terminal to make sure
you are located in your home directory:
cd
2. Type the following line of code into your Terminal and press
Enter to navigate to the folder that the EWS is locate in:
cd downloads/osp/Embedded\ Webserver
3. Type the following line of code into your Terminal and press
Enter to get Composer
brew install composer
4. Type the following line of code into your Terminal and press
Enter to install Composer
composer install
5. Type the following line of code into your Terminal and press
Enter to get NPM
brew install npm
6. Type the following line of code into your Terminal and press
Enter to install PHP
brew install php
7. Type the following line of code into your Terminal and press
Enter
cp .env.example .env
24
8. Type the following line of code into your Terminal and press
Enter
php artisan key:generate
9. Type the following line of code into your Terminal and press
Enter to install React
npm install react
10. Type the following line of code into your Terminal and press
Enter to install React-dom
npm install react-dom
11. Type the following line of code into your Terminal and press
Enter to install NPM
npm install
12. Type the following line of code into your Terminal and press
Enter to run NPM
npm run dev
5.1
Running RT-MHA on Local Computer
1. Enter the following line of code In the same Terminal Window
in which you ended the previous step (npm run dev), type in the
following line of code and press Enter
php artisan serve
2. Press the ’Run’ Triangle in the Upper left hand corner of your
XCode window to run the OSP code
3. Open a new Terminal Window Type in the following line of code
and press enter:
php artisan socket:start
5.2
Connecting to OSP Remotely
1. The following section is specifically for running OSP on a Mac and
connecting to it remotely.
25
2. The first thing that you should do is download a text editing
software such as Atom or Sublime To download, go to this link
and press download: https://atom.io/
3. When it finishes downloading double click on the downloaded
item, follow the instructions to move it into your Applications
folder, and open a new window in Atom when it is finally loaded.
4. Open a new Finder Window and double click on your Hard
Drive to see the contents stored there.
Figure 5.9: "This is what your Finder
window should look like when you
have selected your Hard Drive
5. Press following three together: "command"+"shift"+">" Once you
do this, you should see some grayed out folders appear in the
Finder window, that were not there before
Figure 5.10: Here are all the files you
have made visible with this command,
with the etc folder highlighted
26
6. Double click the grayed out folder "etc", then go into the folder
titled "Apache", and find the file titled "httpd.conf"
Figure 5.11: Here are all the files you
have made visible with this command,
with the etc folder highlighted
7. Right click on this httpd.conf file and select Open With> Atom
To open the file in your text editor. When you do this, it will open
the contents of this file for you to edit using Atom.
Figure 5.12: This is how you Open the
httpd.conf file using Atom
8. Once you’re looking at the file in Atom, and can see all of the
text, press the following buttons together "Command"+"f" and
then search for the text in the line below:
27
# LoadModule php7_module l i b e x e c / a p a c h e 2 / l i b p h p 7 . s
When you get to this line, delete the hashtag (#) at the beginning
of the line to complete this step
Figure 5.13: This is what it should
look like after you delete the # on the
appropriate line
9. Press "Command"+"f" and then search for the text in the line
below:Once you find the line, delete the hashtag (#) at the beginning of the line to complete this step
#LoadModule rewrite_module libexec/apache2/mod_rewrite.so
Figure 5.14: This is what it should
look like after you delete the # on the
appropriate line
10. Open a new terminal window and type in the command below
Copy or remember whatever the output of this command was
28
for the next step. For your reference in the images, my mine is
tzubatiy.
whoami
Figure 5.15: This is what your Terminal
window will look like when you use
the whoami command. The output is
highlighted. Your output will be your
username, not tzubatiy.
11. Go back into your Atom window, press "command"+"f" and then
search and then search for the text in the line below: when you
find it, replace the _www with the ouptput of step 9
User _www
Figure 5.16: This is what your Atom
window should look like when have
added your name to the user line, in
this case this is my name.
12. On the line after "Group _www", copy in the following code:
ServerName localhost
DocumentRoot "home/downloads/OSP/embedded\ Webserver/public"
AllowOverride All
Options Indexes MultiViews FollowSymLinks
Require all granted
Figure 5.17: This is what your Atom
window should look like when have
added these lines of code in the right
place.
13. In your Atom window press the buttons "Command"+"s" to save
the changes you have made, and close the file.
29
14. Type the following line of code into a new Terminal window
and press Enter:
sudo apachectl restart
15. Type the following line of code into a new Terminal window
and press Enter:
sudo chgrp -R www-data __PATH__/embeddedwebserver/
16. Type the following line of code into a new Terminal window
and press Enter:
sudo chmod -R 775 __PATH__/embeddedwebserver/storage/
17. Type the following line of code into a new Terminal window
and press Enter:
sudo chmod -R 775 __PATH__/embeddedwebserver/bootstrap/
18. Type the following line of code into a new Terminal window
and press Enter:
sudo apachectl restart
19. To connect to the OSP and control the RT-MHA REMOTELY, use
the extension -t
This starts the application in TCP daemon mode. In this mode, the
application is waiting for a connection on port 8001 from an OSP
client. (See section 3.2) Once a client connects, RT-MHA can be
controlled remotely, such as the EWS, controlled using the web
apps described in section 4.
To launch OSP in daemon mode, type the line below into your
terminal and press Enter:
osp -t -T2 -d-40 -a -r
Notice how in the terminal window it now says:"Using TCP in
daemon mode" and "waiting for connection from client..."
20. Open a new terminal window, copy in the following line of code
and press Enter:
cd
21. Type the following line of code into the Terminal window from
step 20 and press Enter:
cd Downloads/OSP/Embedded\ Webserver
30
22. Type the following line of code into this same Terminal window
and press Enter:
php artisan socket:start
23. Confirm that you are connected to the same wireless network on
this Mac, as on the device you are trying to connect with
24. Open a browser window on the device you are trying to connect
with, and enter your IP address as the URL to start
Verifying the EWS is Working
For verifying that the system is working properly, we recommend
using Apple earbuds. Put them into your computer before adjusting
volume in step 2 below.
6.1
Researcher Page
1. If you still have OSP running, press "command"+"c" Otherwise,
proceed to the next step
2. Open a new terminal window, copy in the following line of code
and press Enter
osp -t -T2 -d-40 -a -r
3. Copy the following line of code in the same Terminal Window
in which you ended the previous step, and press Enter
php artisan serve
4. Open a new Terminal Window, copy in the line below and press
Enter:
cd Downloads/OSP/Embedded\ Webserver
5. Type in the following line of code into the Terminal window
from the previous step (4.1.4) and press Enter:
php artisan socket:start
6. Open your browser, and go the following URL
http://localhost:8000/researcherpage
7. Increase the volume on your computer to the maximum
8. Add at least 5db to each cell in the the G65 row
32
9. Press transmit and listen to the sound that plays
10. Make another change, let’s say removing those 5db and listen
again
11. As soon as you have heard a difference between two parameter
states, you can be sure that the software is working as intended.
6.2
4AFC
1. If you still have OSP running, press "command"+"c" Otherwise,
proceed to the next step
2. Open a new terminal window, copy in the following line of code
and press Enter
osp -t -T2 -d-40 -a -r
3. Copy the following line of code in the same Terminal Window
in which you ended the previous step, and press Enter
php artisan serve
4. Open a new Terminal Window, copy in the line below and press
Enter:
cd Downloads/OSP/Embedded\ Webserver
5. Type in the following line of code into the Terminal window
from the previous step (4.2.4) and press Enter:
php artisan socket:start
6. Open your browser, and go the following URL
http://localhost:8000/4afc
7. Click on the four choices and the play symbol and make sure
you are hearing the appropriate word
8. Once you hear the words playing on click, you know that the
4AFC app is working properly
Bibliography
1. Harinath Garudadri, Arthur Boothroyd, Ching-Hua Lee, Swaroop
Gadiyaram, Justyn Bell, Dhiman Sengupta, Sean Hamilton, Krishna Chaithanya Vastare, Rajesh Gupta, and Bhaskar D. Rao, "A
Realtime, Open-Source Speech-Processing Platform for Research in
Hearing Loss Compensation", In Signals, Systems and Computers,
2017 51st Asilomar Conference, IEEE, 2017.
2. James M. Kates, "Principles of digital dynamic-range compression," Trends in Amplification, vol. 9, no.2, pp. 45-76, 2005.
3. Shilpi Banerjee, The compression handbook: An overview of the
characteristics and applications of compression amplification, 4th
Edition, Starkey Laboratories, 2017.
4. James M. Kates, Digital hearing aids, Plural publishing, 2008.
5. Toon van Waterschoot and Marc Moonen, "Fifty years of acoustic
feedback control: State of the art and future challenges," Proceedings of IEEE, vol. 99, no. 2, pp. 288-327, 2011.
6. Johan Hellgren, "Analysis of feedback cancellation in hearing
aids with Filtered-x LMS and the direct method of closed loop
identification," IEEE Transactions on Speech and Audio Processing,
vol. 10, no. 2, pp. 119-131, 2002.
7. Hsiang-Feng Chi, Shawn X. Gao, Sigfrid D. Soli, and Abeer Alwan,
"Band-limited feedback cancellation with a modified filtered-X
LMS algorithm for hearing aids," Speech Communication, vol. 39,
no. 1-2, pp. 147-161, 2003.
8. Donald L. Duttweiler, "Proportionate normalized least-meansquares adaptation in echo cancelers," IEEE Transactions on
Speech and Audio Processing, vol. 8, no. 5, pp. 508-518, 2000.
9. Jacob Benesty and Steven L. Gay, âĂIJAn improved PNLMS algorithm,âĂİ in Proceedings of IEEE International Conference on
Acoustic, Speech, and Signal Processing (ICASSP), pp. 1881-1884,
2002.
34
10. Constantin Paleologu, Jacob Benesty, and Silviu CiochinÄČ, "An
improved proportionate NLMS algorithm based on the l0 norm,"
in Proceedings of IEEE International Conference on Acoustic,
Speech, and Signal Processing (ICASSP), pp. 309-312, 2010.
11. Arslan, Levent, Alan McCree, and Vishu Viswanathan. "New
methods for adaptive noise suppression." Acoustics, Speech, and
Signal Processing, 1995. ICASSP-95., 1995 International Conference
on. Vol. 1. IEEE, 1995.
12. Hirsch, Hans-GÃijnter, and Christoph Ehrlicher. "Noise estimation
techniques for robust speech recognition." Acoustics, Speech, and
Signal Processing, 1995. ICASSP-95., 1995 International Conference
on. Vol. 1. IEEE, 1995.
13. Cohen, Israel, and Baruch Berdugo. "Noise estimation by minima
controlled recursive averaging for robust speech enhancement."
IEEE signal processing letters 9.1 (2002): 12-15.
14. Lee, Ching-Hua, Bhaskar D. Rao, and Harinath Garudadri. "Sparsity promoting LMS for adaptive feedback cancellation". In Signal
Processing Conference (EUSIPCO), 2017 25th European, pp. 226230. IEEE, 2017.
Source Exif Data:
File Type : PDF
File Type Extension : pdf
MIME Type : application/pdf
PDF Version : 1.5
Linearized : Yes
Author : OSP Team
Create Date : 2019:06:19 00:02:12Z
Creator : LaTeX with hyperref package
Modify Date : 2019:06:19 00:02:12Z
PTEX Fullbanner : This is pdfTeX, Version 3.14159265-2.6-1.40.18 (TeX Live 2017) kpathsea version 6.2.3
Producer : pdfTeX-1.40.18
Subject :
Title : Open Speech Platform User guide
Trapped : False
Page Mode : UseOutlines
Page Count : 34
EXIF Metadata provided by EXIF.tools
.