ARTIFICIAL VOICE SYNTHESIZER

USING FUZZY LOGICS

THE TRANSLATOR FOR DUMB

1
Dr.Adiline Macriga.G2Kruthiga.K3Anusudha.S
1
Professor 2Student 3Student
123
Department of Information Technology
123
Sri Sairam Engineering College, Chennai-600044, India.
1
[email protected] [email protected]@gmail.com

Abstract: voice of the dumb people) as output which

makes others to understand their comments
Speech is one the most important easily. Thus the “Artificial voice synthesizer”
ways of interaction in this drastically comfort the dumb person in overcoming their
developing communication era. Lamentably, disability by communicating like a normal
the mute people are dispossessed of this boon. person.
“Artificial Voice synthesizer” helps them to
overcome this difficulty and communicate Keywords— neural impulses- processing of
with the world in an efficient way. The signals - conversion of signals into voice
principle objective of the artificial voice messages- Artificial intelligence- fuzzy logics.
synthesizer is to make dumb people to
A. Introduction:
communicate easily with other people even
In the research of “Artificial voice
who cannot understand the sign language.
synthesizer” the main innovative concept used is
Communication can be made easier by giving
the neural impulse sensing and the artificial
artificial voice to the dumb people so that
intelligence which is the next era of technology.
they do not face any kind of difficulty in
Artificial intelligence is the ideology and
communicating with the world.This device
augmentation of computer systems that are able
makes use of neural sensors which is fixed
to perform tasks normally requiring human
over the head and behind the user’s ear lobe.
intelligence, such as visual conception, speech
The users can make their usual signs for
identification, decision-making, and translation
conveying their ideas. The sensors senses the
between languages. Usage of brain impulse is
brain impulses as each hand movement have
even though complex to understand, it can make
a specific set of brain impulses and then some
world dynamically change according to users’
specific calculations are made using Discrete
wish when used properly. Every time when a
Fourier Transform. The fuzzy logics in
mute person wants to makes a gesture few nerve
Artificial intelligence is used to match the
impulses are generated which are transmitted to
obtained solutions of the converted signals
the brain using acetylcholine that is secreted by
with the stored solutions so that it converts
the neurons. The brain again sends back a
the impulses into voice and gives that
response in a similar way to the hand to perform
voice(the translated, artificially synthesized
the gesture. The response of the brain is
978-1-5090-6221-8/17/$31.00 2017
c IEEE 388
collected as the brain impulse and made use to TI’s MSP430 microcontroller.
generate voice. The process that takes place in
the system can be categorized into steps and
described as follows.
B. Extraction of brain impulses:
Naturally human brain responds by
generating specific impulses for each and every
action we perform. More than 2,000,000 neurons
coordinate to perform single action. Action
potentials in nerve cells are also known as
"nerve impulses" or "spikes", and the secular
sequence of action potentials generated by a
Figure 2
neuron is called its "spike train". The impulses
are transmitted through a chemical substance
called acetylcholine. There are several hand The MSP430 can be used for low
gestures in the sign language as shown in the powered embedded devices which can be used
Figure 1 which have specific brain impulses. A as wearables as the current drawn in idle mode
gesture is a form of nonverbal communication can be less than 1 microampere. The top CPU
or non vocal communication in which visible speed is 25 MHz. In addition to these features
body actions are used to communicate particular the MSP430 is capable of wakeup times below 1
messages, either in place of, or in juxtaposition microsecond, allowing the microcontroller to
with speech. These hand mimes are used for stay in sleep mode longer, reducing its average
indicating specific words. Mute people current consumption. The device comes in many
communicate with others using this sign configurations consisting of the usual
language. peripherals like internal oscillator, timer
including Pulse Width Modulation(PWM),
Universal Synchronous/Asynchronous
receiver/Transmitter(USART), watchdog,
10/12/14/16/24-bit ADCs, Serial/Parallel
Interface (SPI), I²C and brownout reset circuitry.
Some of the very rare peripheral options include
comparators that can be used with the timers to
do simple ADC are used. The comparators are
on chip op-amps for signal conditioning, LCD
Figure 1 driver,12-bit DAC, hardware multiplier,
Universal Serial Bus (USB), and Direct Memory
Access (DMA) for ADC results. Thus the
The Figure 2 is to explain that there are MSP430 is much more advantageous than other
specific brain impulses for specific human low power consuming controllers.
activity.
C. Processing of the signals:
The main aim is to extract the specific nerve Extracted impulse using an EEG sensor or an
impulse that is generated by the brain for the electrode is used as the input for MSP430 and
hand movements that a dumb person performs the signals are converted into digital signal using
for responding to people. This is achieved by

2017 Second International Conference On Computing and Communications Technologies(ICCCT’17) 389

AD9680 (Analog to Digital Converter). The obtaining the particular solution. These solutions
analog signals are converted as because they are mapped with their respective textual words.
cannot be processed directly and converted into The mapping of fuzzy logic with their respective
voice. The obtained digital signal is transmitted words is done using Artificial Intelligence.
to a handheld device which is specially designed Prediction of most commonly used words and
for the “Artificial voice synthesizer” using words cache are done using artificial
wireless transmitter. This is diagrammatically intelligence. The words with their specific fuzzy
explained in Figure3. solution are stored in the database. The fuzzy
solutions are compared with the stored value and
the specific words are recognized. The
comparison of the signal is done as given in the
Figure 4.

Figure 3

In the hand held device the solutions Figure 4

which are obtained by transforming the fuzzy
logics of impulse generated for each action of
the hand movement are stored. Fuzzy logic is a D. Conversion of signals into
form of many valued logic in which the truth voice:
values of variables may be any real number The texts that are stored are also in the
between 0 and 1 unlike in Boolean logic, the form of binary data. Hence these data are to be
truth values of variables may only be the integer again converted into analog signals as they have
values 0 or 1. Fuzzy logic has been employed to to be given out as voice signals. To convert the
handle the concept of partial truth (brain digital data into analog signals a Digital to
impulses are always non uniform) , where the Analog Converter (Digital Optical Coaxial to
truth value may range between completely true slink to Analog RCA Audio Converter Adapter)
and completely false. In Artificial Voice is used. A digital to analog converter is a device
Synthesizer the fuzzy logics is used because we that is capable of converting a digital signal into
use brain impulse input which are always non- an analog signal. An ideal Digital to Analog
uniform. Henceforth we cannot obtain the exact Converter converts the discrete numbers or any
solution for the transformed brain impulse. The discrete signal into a conceptual continuous
brain impulses are transformed using specific sequence of impulses that are then processed by
transformation schemas and the solution is a reconstruction filter using some form of
obtained. The transformation includes the discursion and sampling to fill in data between
application of transforming functions like the impulses. A conventional practical Digital to
Fourier transform and discrete transform and Analog converter converts the numbers into a
hybrid constant function made up of a sequence

390 2017 Second International Conference On Computing and Communications Technologies(ICCCT’17)

of rectangular functions that is modelled with
the zero-order hold. The analog signal that is
obtained from Digital to Analog converted is
converted into voice using Visaton EX60s piezo
transducer 8 ohm (Voice transducer) and given
out using a micro speaker. Audio Sound
Transducers include both input sensors that
convert sound into an electrical signal such
as a microphone, and output actuators that
convert the electrical signals back into sound
such as a loudspeaker. This is explained in
Figure 5 where digital data is converted into
voice signals. Figure 6

The Figure 7 helps us to understand the

system architecture of “Artificial Voice
Synthesizer” where the user is wearing the
electrodes for receiving the brain impulse. The
obtained brain impulse is processed and voice
signal is converted into voice signal. In the
Figure 6 analog brain impulses are converted
into digital signals and then transmitted where it
is converted into appropriate voice message.

Figure 5
F. System Architecture

E. System Flow
The system flow can be explained as in
the Figure 6. Initially the user performs hand
movement which generates a specific brain
impulse. These impulses are converted into
digital signals using analog to digital
converter. The converted signal is
transmitted to the handheld device where it
is converted into fuzzy logics. These fuzzy
logics are transformed and appropriate Figure 7
solutions are obtained which is compared
with already stored solutions (the solution of
the signal stored in the database). The G. Future Scope
textual representation of the compared In the proposed system only the hand
output is obtained and converted into voice gestures are considered for generating the voice
signal using the voice transducers. message. According to the study of gesture

2017 Second International Conference On Computing and Communications Technologies(ICCCT’17) 391

language facial expression also play a major role Applications", ISAST
in communication. Thus in addition to the Transactions on Computers and
evaluation of hand gestures the facial expression Intelligent Systems, vol. 1(2),
can also be combined and evaluated. This can 2009
make the device work more accurately. The [6] “Virtual Voice” --- The voice
facial expression can be tracked with the help of for the dumb K. Rajeswari; E.
the facial muscle movement using any tracking Jeevitha; V. K. G. Kalai
systems like infrared tracking. This may help in Selvi.
making the communication in a better way. The http://ieeexplore.ieee.org/docu
database can be improvised by updating it with ment/5705909/
the desired languages and new gestures. This
[7] Continuous gesture recognition
can also improvise the communication using the
by using gesture spotting Daeha
“Artificial voice Synthesizer”.
Lee; Hosub Yoon; Jaehong Kim
2016, 16th International
H. References: Conference on Control,
[1] Augarten, Stan (1983). The Automation and Systems
Most Widely Used Computer on (ICCAS)
a Chip: The TMS 1000. State of http://ieeexplore.ieee.org/docu
the Art: A Photographic History ment/7832502/
of the Integrated Circuit (New [8] MSP430 Mixed Signal
Haven and New York: Ticknor Microcontroller description
& Fields). ISBN 0-89919-195- http://www.ti.com/lit/ds/symli
9.Retrieved 2009-12-23. nk/msp430f1611.pdf
[2] David Harris & Sarah Harris
(2012). Digital Design and
Computer Architecture, Second
Edition, p. 515. Morgan
Kaufmann. ISBN 0123944244.
[3] Momentum Carries MCUs Into
2011 http://semico.com/content/
momentum-carries-mcus-2011
[4] Edwards, Robert
(1987). "Optimizing the Zilog
Z8 Forth Microcontroller for
Rapid Prototyping" (PDF).
Martin Marietta.
p. 3.Retrieved 9
December 2012.
[5] D. Peters, D. Raskovic, and D.
Thorsen, "An Energy Efficient
Parallel Embedded System for
SmallSatellite
392 2017 Second International Conference On Computing and Communications Technologies(ICCCT’17)