Вы находитесь на странице: 1из 4

2009 International Conference on Computer Technology and Development

Homes Appliances Controlled Using Speech Recognition in Wireless Network


Environment

Mardiana B., Hazura H.,Fauziyah S., Zahariah M.,Hanim A.R., Noor Shahida M.K.
Faculty of Electronics and Computer Engineering,
Universiti Teknikal Malaysia Melaka
Ayer Keroh, Malacca, Malaysia
mardiana@utem.edu.my,hazura@utem.edu.my,fauziyah@utem.edu.my,zahariah@utem.edu.my,zahariah@utem.edu.my,
hanim@utem.edu.my, noorshahida@utem.edu.my

Abstract— This paper highlights the development of a human system. This system can help to make the daily chores at
voice command system that is able to activate the electrical home much easier for disabilities user.
appliances at home in wireless environment. The idea is to
establish a wireless connection between a client computer and a II. METHODOLOGY
server in a real application. It is developed especially to assist
disabled person to do their house chores. Via the wireless
communication system, this system utilizes the remote voice
This system is applied between two computers (a
recognition system server to translate the voice input received server and a client) that perform in a real time
from a serviced client computer into a symbolic data file to be communication system. The software part consists of three
processed by the client’s computer. The wireless technology main parts, the server computer, the client computer and the
implemented in this project is the Wi-Fi network and the wireless network interfacing. The wireless communication
computer server utilizes the GUI manager for implementing the system utilizes a remote voice recognition server system to
speech recognition system. The GUI system is developed using translate voice input received from serviced client computer
the Microsoft Visual Basic.net software and the speech into a symbolic data file (e.g. alpha-numeric or control
recognition software used is the Microsoft Speech (SAPI). In
characters) that can be processed by the client computer.
conclusion, from the result analyses, the performance of the
system depends on the wireless coverage and distance, voice
The translation process begins by establishing a wireless
input type and the environment condition. voice communication channel between the client computer
and the voice recognition server.
Keywords- speech recognition; wireless network; home A user of the client computer then begins speaking in a
appliances fashion that may be detected by the voice recognition server
system. Upon detecting the user's speech, the voice
I. INTRODUCTION recognition server system translates the speech into a
symbolic data file, which is then forwarded to the user
Speech Recognition is a technology which allows the through a separate data communication channel. GUI
computer to identify and understand words spoken by a command will send data to the interface circuit through the
person using a microphone or telephone. The speech that parallel port. The interface circuit will trigger the interface
user and computer exchanges are scripted. In other words, circuit and activate the electrical appliances and load as
user can talk with computer using a set of pre-programmed required by the user.
commands and instructions. The computer will respond in This paper discusses the methodology of developing of
the same way (also using a scripted language) [6]. Computer software part and setting up the wireless network interfacing.
software that understands input speech enables user to have Fig.1 shows the overall system architecture.
conversations with the computer. These conversations would
include user and the computer speaking as commands or in A. Software Development
response to events, input, or other feedback. Speaking is The software part of this system is includes the GUI
easier and more intuitive than selecting buttons and menu development and speech recognition development. The GUI
items. Human speech has evolved over many thousands of is developed using Microsoft Visual Basic.Net and the
years to become an efficient method of sharing information speech recognition software utilizes the Microsoft Speech
and giving instructions. (SAPI)[1]. Before the system operates, it is necessary to
The application of the speech recognition technology has declare the speech recognition engine in the same with form
been implemented in this project. This project is to produce a Voice Recon-frmTest (Form). The declaration must be done
system that can control any type of electrical appliances in to ensure that the system is ready to be run and can be linked
the house such as lamp, fan, radio, television, etc by using with other systems. Besides that, Microsoft Speech must be
the user’s voice command. User’s voice command is an included in reference to the whole coding development
input to activate the electrical appliances through this

978-0-7695-3892-1/09 $26.00 © 2009 IEEE 285


DOI 10.1109/ICCTD.2009.178
[5].Fig. 2 depicts the GUI main page that has been developed
in this project.

B. Setting Up Wireless Interfacing


The steps in creating a wireless network are selection of
wireless equipment, connection of wireless router and
configuration and finally, the connection to computers or
laptop.

• Application Networking: To ensure the system


operates as desired, all parts must be connected in
wireless system. The GUI for client system must
support linkages between speech recognition system
and the hardware engine. At the server side, the
system also must be programmed to accept the voice
data from wireless medium. For wireless
networking, it is compulsory to set-up in area that
has signal coverage [2].
• Remote Desktop: Remote Desktop allows user to
Figure 1. Overview of system architecture
have access to a Windows session that is running on
his/her computer, when the user is at another
computer. In order to perform share application, both
command is given. Five types of devices are being tested
computers should run the same operating system i.e.
the Microsoft Windows XP. Besides that, both client including Lamps, Fans, Air-conditioner, TV and Radio. The
and user must be connected via Windows Messenger changes of button color indicate the level of response and
and they need to set-up an account for sign-in action from each appliance. Fig. 3, 4 and 5 show the result
purpose. After being connected, both clients and user of appliance system activation for certain commands.
will be able to view computer screen and chat online Results show that the GUI Appliance System Activation is
together in real time. With permission, both clients able to respond to the input command accordingly.
and user can even use their mouse and keyboard to This system needs wireless network to operate and
communicate with others computer. Remote the performance is based on type and range of the coverage.
Desktop also provides a Fast User Switching, which The parameter that must be considered for wireless
can easily switch from one user to another on the coverage analysis is the performance based on the distance
same computer. or wireless coverage.
• Security by Windows Firewall:For security purpose, Fig. 6 shows that the performance of the system is
Windows Messenger needs to have permission from inversely proportional with the range or distance of wireless
Windows Firewall to connect to the Internet to use coverage. When a client communicates with server from a
features such as file transfer, voice and video, short distance, the servers can respond quickly and vice
Application Sharing, and Whiteboard. This versa. Besides that, the coverage distance is limited by the
permission is established by using the Exceptions list type of wireless router that being used.
in Windows Firewall. In Microsoft Windows XP
Service Pack 2 (SP2), Windows Messenger is
automatically added to the Windows Firewall
Exceptions list the first time user log on to .NET
Messenger Service. User must be logged in as an
administrator or a member of the Administrators
group in order to complete this procedure. If the
computer is connected to a network, network policy
settings might also prevent you from completing this
procedure.
III. ANALYSIS AND DISCUSSION
A few analyses have been done to evaluate the performance
of the system.
A. Analysis of Appliance System Activation
This analysis is done to measure the capability for Figure 2. GUI main display
each home appliance device to response after certain

286
below. For Microsoft Mike and Microsoft Sam, this text to
speech recognition have been programmed and installed as
package with the Window XP. Both Microsoft Mike and
Microsoft Sam only suitable for man and the user can switch
the type of voice, depending on their voice tone.
For Microsoft Mary, these types of text to speech
recognition also have been programmed in the Window XP
software. This setting can be done at the control panel
window. Microsoft Mary is suitable for ladies or woman.
Sample TTS Voice needs a collection of some samples of
user voice so that it can be adjust to the user speaking style.
This method needs a special time to train and collect the
speech style and must be repeated for a better recognition.
For these input voice, the performance for voice recognition
is more accurate if the training is perfectly done. The input
setting is assigned for only one user and the dictation and
recognition quality suffers if other users use the systems.
Figure 3. Condition - All appliances OFF
From Fig. 7, it shows that Microsoft Mike is the best
performance for input recognition and dictation. In general,
speech or voice interfacing can be implemented in many
applications such as the home appliances control, computer
application control and kiosk machine. The advantage of this
system is the mobility of the user which can access a server
from any location which has wireless coverage. Besides that,
the input voice is friendlier for any stage of ages and gender
because it can be set-up manually by the user. The
interfacing and link between PC to PC is more convenience
because both user and clients can have a full view of each PC
desktop. This allows easier control and management of the
home appliances system besides performing other tasks. The
connection using the Windows Messenger is more secure
since user needs to log-on for security purpose. The systems
can only be control remotely if both client and the user have
an agreement or permission. Nevertheless, the system also
can be activated using the button if the voice recognition
Figure 4. Condition - All appliances ON system is down. This system can be handled personally and
the development cost is still competitive and it meets client
satisfactions.

B. Analysis of Wireless Coverage


This system needs wireless network to operate and
the performance is based on type and range of the coverage.
The parameter that must be considered for wireless
coverage analysis is the performance based on the distance
or wireless coverage.
Fig. 6 shows that the performance of the system is
inversely proportional with the range or distance of wireless
coverage. When a client communicates with server from a
short distance, the servers can respond quickly and vice
versa. Besides that, the coverage distance is limited by the
type of wireless router that being used.

C. Analysis of Input Voice


The type of input voices to be recognized is fundamental
to this system. The type of input voice can be set up at the Figure 5. Performance of Interfacing over Distance
window control panel, under the speech properties. For this
analysis, three types of input voice are tested as shown

287
PC is more convenience because both user and clients can
have a full view of each PC desktop. This allows easier
control and management of the home appliances system
besides performing other tasks. The connection using the
Windows Messenger is more secure since user needs to log-
on for security purpose. The systems can only be control
remotely if both client and the user have an agreement or
permission. Nevertheless, the system also can be activated
using the button if the voice recognition system is down.
This system can be handled personally and the development
cost is still competitive and it meets client satisfactions.
In conclusion, this system recognizes the input
Figure 6. Performance of Interfacing over Distance
commands very well and manages to give a good response if
the wireless coverage have strong signal in order to get a
good interfacing result. The type of input voice that suitable
for this is Microsoft Mike. Another consideration is the
system must run in a clear and quiet environment condition.
Other disruption such as weather, buildings and noise must
be avoided to get a better performance. The wireless
connection will be slightly slower than the direct connection
because a Wi-Fi transmits through the air and there's some
degradation with this medium.

ACKNOWLEDGMENT
The authors would like to thank UTeM as the sponsor,
Dean, Assoc. Prof. Muhammad Syahrir Bin Johal and all
Head of Departments in the faculty.
Figure 7. Input Speech Dictation

REFERENCES
In conclusion, this system recognizes the input
[1] Julia Case Bradley, Anita C. Millspaugh. Programming in
commands very well and manages to give a good response if Visual Basic. Net. 4th edition. McGraw-Hill Companies, Inc,
the wireless coverage have strong signal in order to get a New York: McGraw-Hill.4-6; 2003
good interfacing result. The type of input voice that suitable [2] Andy Wigley, Peter Rexburg. Building .Net Applications for
for this is Microsoft Mike. Another consideration is the Mobile Devices. Washington: Microsoft Press. 35-37; 2002.
system must run in a clear and quiet environment condition. [3] Martyn Mallick. Mobile and Wireless Design Essentials.
Other disruption such as weather, buildings and noise must Indianapolis USA: Wiley Publishing, Inc ; 153-193.2003.
be avoided to get a better performance. The wireless [4] Roy Blake. Wireless Communication Technology. USA:
connection will be slightly slower than the direct connection Delmar Thompson Learning; 1-100. 2001.
because a Wi-Fi transmits through the air and there's some [5] Soo-Young Suk.Ho-Youl Jung. Shozo Makino. Hyun-Yeol
degradation with this medium. Chung. Distribution Speech Recognition System for PDA in
Wireless Network Environment. SPECOM’2004:9th
Conference Speech and Computer. ISCA Archive.
September 2004.
IV. ANALYSIS AND DISCUSSION [6] James R Mault. Speech Recognition Capability for a
In general, speech or voice interfacing can be Personal Digital Assistant. US2003/0163321 A1. August
implemented in many applications such as the home 2003.
appliances control, computer application control and kiosk [7] Azuraini Binti Muda. Membangunkan Sistem Perisian Untuk
Kawalan Litar Elektrik Menggunakan Suara. Tesis Sarjana
machine. The advantage of this system is the mobility of the Muda. Universiti Teknikal Malaysia Melaka; 2007
user which can access a server from any location which has
wireless coverage. Besides that, the input voice is friendlier
for any stage of ages and gender because it can be set-up
manually by the user. The interfacing and link between PC to

288

Вам также может понравиться