Skip to content
Activate installed microphone, and analyze sound input
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.

README.md

LeQue

The intention of this web-application is to allow the visitor to provide auditory commands, which get executed on the server. The languages used for this project includes:

  • PhP
  • Python
  • Bash-Script
  • JavaScript
  • Sass (Ruby)

Note: Since this project is open to the public, feel free to contribute, or email jeff1evesque@yahoo.com regarding any questions.

###Definition

Speech recognition (SR) is the translation of spoken words into text.

###Overview

This project utilizes Flash within the web-browser in order to access the users microphone. When a recording is saved, it is reconfigured to a 16 bit, 16 kHz, mono if it has a different format. This reconfigured wav file, is then converted to text using PocketSphinx, which allows our Python scripts to parse the converted text into executable commands. This Bash Automation occurs automatically, and triggered when an audio file is saved either through the web-application, or by manually saving a wav file into the recording/ subdirectory.

##Installation

###Ubuntu Server 14.04

Format two USB flash drives as MS-DOS (FAT). Using UNetbootin, make both USB drives bootable with the following ISO images:

Next, ensure the machine being used has been partitioned enough unallocated space for the Ubuntu Server 14.04 operating system.

Note: Current development is on a windows machine, hence formatting the latter USB drives to MS-DOS (FAT).

Note: The unallocated space on the hard-disk does not need to be formatted before installation.

During installation (boot-up with the Ubuntu USB), select Install. At the [!] Partition disks section, select Manual, and carefully partition only the unallocated space on the hard-disk. This is important if the overall machine is a multiboot. The partitioning process will create two partitions, Ext4, and SWAP from the unallocated space. Upon reaching the [!] Software Selection section, select Basic Ubuntu Server, and Ubuntu desktop. Also, select yes when asked Install the GRUB boot loader to the master boot record?

Note: If Ubuntu Server 14.04 was not bootable on the hard disk after installation, use the boot-repair-disk bootable USB, and reinstall Grub.

###Linux Packages

The following packages need to be installed through terminal in Ubuntu:

# General Packages:
sudo apt-get update
sudo apt-get install inotify-tools
sudo apt-get install libav-tools
sudo apt-get install firefox
sudo apt-get install flashplugin-installer
sudo apt-get install git-core
sudo apt-get install lamp-server^ phpmyadmin

# Sphinx Packages: allow `./autogen.sh`, `sudo make install` for submodules
sudo apt-get install autoconf
sudo apt-get install libtool
sudo apt-get install bison
sudo apt-get install swig
sudo apt-get install python-dev

##Configuration

###GIT

Since we installed GIT earlier, we have to remember to configure our GIT user. Only change the values within the double quotes (remove the quotes for the email):

git config --global user.email "YOUR-EMAIL@DOMAIN.COM"
git config --global user.name "YOUR-NAME"

Fork this project in your GitHub account, then clone your repository:

cd /var/www/html/
sudo git clone https://[YOUR-USERNAME]@github.com/[YOUR-USERNAME]/leque.git [PROJECT-NAME]

Then, add the Remote Upstream, this way we can pull any merged pull-requests:

cd /var/www/html/leque/
git remote add upstream https://github.com/[YOUR-USERNAME]/[REPOSITORY-NAME].git

####GIT Submodule

We need to initialize our git submodules:

sudo git submodule init
sudo git submodule update

Note: We have to use the sudo prefix, since we haven't taken care of file permissions yet.

The above two commands will update submodules. If they are already initialized, then the latter command will suffice. Then, we need to pull the code-base into the initialized submodule directory:

cd /var/www/html/leque/
git checkout -b NEW_BRANCH master
cd [YOUR_SUBMODULE]/
git checkout master
git pull
cd ../
git status

Now, commit and merge the submodule changes.

###File Permission

Change the file permission for the entire project by issuing the command:

cd /var/www/html/
sudo chown -R jeffrey:sudo leque

Note: change 'jeffrey' to the user account YOU use.

Then, with the exception of the .gitignore file, ensure /var/www/html/leque/audio is an empty directory, so that we can change it's ownership:

cd /var/www/html/leque/
sudo chown www-data:sudo audio

###PocketSphinx

# Install Sphinx Engine(s)
cd /var/www/html/leque/pocketsphinx/sphinxbase/
./autogen.sh
sudo make install

cd ../pocketsphinx/
./autogen.sh
sudo make install

cd ../sphinxtrain/
./autogen.sh
sudo make install

# Extract Sphinx Acoustic, and Language Models
cd ../../
wget http://sourceforge.net/projects/cmusphinx/files/Acoustic%20and%20Language%20Models/US%20English%20Generic%20Acoustic%20Model/en-us.tar.gz/download -O en-us.tar.gz
sudo tar -zxvf en-us.tar.gz -C /usr/local/share/pocketsphinx/model/hmm/
sudo rm en-us.tar.gz

wget http://sourceforge.net/projects/cmusphinx/files/Acoustic%20and%20Language%20Models/US%20English%20Generic%20Language%20Model/cmusphinx-5.0-en-us.lm.dmp/download -O cmusphinx-5.0-en-us.lm.dmp
sudo mv cmusphinx-5.0-en-us.lm.dmp /usr/local/share/pocketsphinx/model/lm/

More information regarding setting-up PocketSphinx, can be found within the README.md file from the jeff1evesque/pocketsphinx repository.

###Autobahn

In order to allow browsers to stream audio to the server, a websocket server is needed. This project uses Autobahn, an open source project that provides WebSocket, and Web Application Messaging Protocols (WAMP) protocol to achieve audio streaming.

More information regarding setting-up Autobahn, and the requirements of streaming audio from the browser to the server can be found within the README.md file from the whisper repository.

###Grunt

We will automate Grunt's task management, which will encompass tools such as Sass, Uglify, Imagemin, and Modernizr. It requires setup only once within each web-application utilizing its tools.

More information regarding setting-up Grunt, can be found within the README.md file from the Grunt repository.

###Bash Scripts

Configuring /etc/rc.local allows bash-scripts to be run during apache2 boot. Since bash_loader loads all the required bash-scripts, all required scripts can be automated by adding the following to /etc/rc.local:

...
# run 'bash_loader' at start-up for '/var/www/html/leque/' application (edited by JL)
cd /var/www/html/leque/bash/ && ./bash_loader > /dev/null 2>&1 &

exit 0

The above configuration may require starting rc.local:

sudo /etc/init.d/rc.local start

###Boot Sequence

This application utilizes GRUB2, a bootloader program, which allows the selection of partition (on the hard disk) to boot from. Modifying the grub configuration file allows the boot sequence to change. This is done by modifying the order of files contained within /etc/grub.d:

$ cd /etc/grub.d/
$ ls
00_header        10_linux      20_memtest86+  30_uefi-firmware  41_custom
05_debian_theme  20_linux_xen  30_os-prober   40_custom         README

Operating systems associated with lower prefixes will be higher in the boot selection sequence. In the case where two partitions exist - Windows 7, and Ubuntu, 30_os-prober will be associated to the Windows 7 partition. Since, linux is prefixed with a lower number, the boot sequence at start-up will list Ubuntu higher in the list, and perhaps default to it during start-up. One way to change this sequence, is to rename 30_os-prober as follows:

$ cd /etc/grub.d/
$ sudo mv 30_os-prober 09_os-prober
$ ls
00_header        09_os-prober  20_linux_xen   30_uefi-firmware  41_custom
05_debian_theme  10_linux      20_memtest86+  40_custom         README
$ sudo update-grub

The last command above, generates the grub configuration file. When the hard-disk is started, we may see the following on the monitor:

GNU GRUB version 2.02^beta2-9ubuntu1

Windows 7 (loader) (on /dev/sda1)
Windows 7 (loader) (on /dev/sda2)
Ubuntu
Advanced options for Ubuntu
Memory test (memtest86+)
Memory test (memtest86+, serial console 115200)

Use the [up arrow] and [down arrow] keys to select which entry is highlighted.
Press enter to boot the selected OS.  `e' to edit the commands
before booting or `c' for a command-line.

Note: Without the above modifications, the Ubuntu option would preceed both Windows 7 options (both point to the same partition).

###Domain Name

Webservers need to define their own server name. Since this project utilizes Apache2 as one of its webservers, the /etc/apache2/apache2.conf file should include the following lines:

...
# Global configuration (edited by JL)
ServerName localhost
...

Providing server identification is useful for various cases (i.e. self-referential redirects), and eliminates repetitive messages when starting, and restarting the server:

$ sudo /etc/init.d/apache2 restart
 * Restarting web server apache2
[Wed Jul 30 08:48:12.303006 2050] [alias:warn] [pid 5457] AH00671: The Alias directive in /etc/phpmyadmin/apache.conf at line 3 will probably never match because it overlaps an earlier Alias.

Note: It is important to remember that apache2.conf will need to be adjusted when the server has a true identity (replacing the defined localhost).

##Testing / Execution

###Test Scripts

Before translating audio files, it is possible to perform a few tests to gauge the PocketSphinx translation engine. For example, the following script tests the command pocketsphinx_continuous against sample.wav file from the pocketsphinx submodule:

cd /var/www/html/leque/bash/tests/
./test_pocketsphinx_continuous

Note: Since the above script uses sample.wav, be sure to initialize all submodules (as outlined in the GIT subsection).

The execution of the above the script will produce a text-file containing the text translation of sample.wav:

cd /var/www/html/leque/audio/recording_text/
pico test_sample.txt

A corresponding log-file is also created:

cd /var/www/html/leque/bash/logs/
pico log_test_pocketsphinx_continuous

###Translation Time

The PocketSphinx translation engine ideally should have a translation time (TR) equal to three times the recording time (RT):

TR = 3 x RT

The translation time (TR) can be verified by checking the output from the command pocketsphinx_continuous. This command will output many lines. However, the ones of particular relevance have a very specific form.

####CPU Time

The CPU Time is the actual execution time for the pocketsphinx_continuous command. Therefore, the sum of all such lines will produce the overall CPU Time for the pocketsphinx_continuous command:

ngram_search_fwdtree.c(xxx): TOTAL fwdxxxx xx.xx CPU x.xxx xRTINFO:

####System Time

The Wall Time is the actual system time for the pocketsphinx_continuous command. A system can pause processes for various operations, including those used in relation to pocketsphinx_continuous. Therefore, possibly a better measure of the overall translation time. The sum of all such lines will produce the overall System Time for the pocketsphinx_continuous command:

ngram_search_fwdtree.c(xxx): TOTAL fwdtxxxx xx.xx wall x.xxx

####Automation Time

If bash automation is being implemented, information pertaining to Translation Time can be acquired from log_bash_loader:

/var/www/html/leque/bash/logs/
pico log_bash_loader

If test_pocketsphinx_continuous was executed:

cd /var/www/html/leque/bash/tests/
./test_pocketsphinx_continuous

then, the translation time information can be found within log_test_pocketsphinx_continuous:

/var/www/html/leque/bash/logs/
pico log_text_pocketsphinx_continuous
You can’t perform that action at this time.