complete seminar report on the topic silent sound technology given by raj niranjan in MCA department of BMS Institute of Technology and Management , avalahalli,bangalore ,karnataka

1. Silent sound technology
Department of MCA 2015-16 Page 1
1. INTRODUCTION
A silent sound technology is a non-speech embedded engine with a variety of lip
movements and facial expression interactive service based on cloud computing.
This technology helps people to communicate in noisy places and to reduce noise
pollution to some extent. It uses lip detection techniques.
Lip detection is a complex problem because of high variability range of lip shapes
and color. This technology aims to analyze and understand-every movement of the
lips and facial expressions then transform them into text and audio output. Humans
are proficient of generating and understanding whispered speech in noiseless
environments at silent strangely low signal levels. Most folks can also understand
little words which are unsaid, by lip reading. With multiple levels of video
processing, it’s possible to obtain lip outline and location of key points in a
subsequent frames is usually referred to as lip tracking. Lip tracking is one of the
biometric systems based on which an unaffected system can be developed.
2. HISTORY
The idea of understanding silent speech by automated means or by means of a
computer has been around for an elongated time, and was promoted in the 1968
Stanley Kubrick science-fiction movie “2001 – A Space Odyssey”.
In that movie with the use of a computer the silent sound was converted into the
amplified audio. A major focus was the DARPA advanced speech encoding
program (ASE) of the initial 2000’s, which financed research on low bit rate
speech synthesis “with acceptable unambiguousness, eminence, and auditory
speaker recognizability in acoustically tough environments”. In the 2000 CeBIT’s
“future park”, a technical fair, concept “silent sound technology” publicized which
targets to notice every single movement of the lips and transfigure them into
sounds. This equipment is being developed by researchers of Karlsruhe Institute of
Technology (KIT), Germany.

2. Silent sound technology
Department of MCA 2015-16 Page 2
3. WHY WE NEED?
When we are in picture show, auditorium, bus, train there is lot of clamor around
us we can’t communicate properly on a mobile phone. It is an end to
uncomfortable situations such as; a person responding his silent cell phone in a
conference, talk or concert, and whispering loudly, “I can’t talk to you right now”.
In the case of a crucial call, remorsefully coming quickly out of the room in order
to reply or call the soul. In future, this problem is abolished with the assistance of
silent sound technology. It is a tool that helps you to transfer information without
using your vocal cords. This equipment notices every lip movements and converts
them into a computer produced sound that can be communicated over a phone.
Hence the person in other end collects the information in audio. It is certainly
going to be a noble solution for those feeling irritated when others speak loud over
the phone.
4. HOW WE SPEAK?
When a human speaks the air passes through the larynx and the tongue and the
words are formed with the help of the articulator muscles in the mouth and the jaw.
The articulator muscles are activated irrespective of the fact that no air passes
through them or not. The weak signals are send from the brain to the speech
muscle. These signals are collectively known as electromyograms.
Fig: -Speaking Process in Human Body [1]

3. Silent sound technology
Department of MCA 2015-16 Page 3
5. METHODS
Silent sound technology is processed in two ways:
5.1 Electromyography
5.2 Image processing
5.1 Electromyography
Electromyography (EMG) involves testing the electrical activities of muscles.
Often, EMG testing is accomplished with another test that measures the conducting
task of nerves. This is termed as nerves conduction study.
Muscular movements comprises the action of muscles and nerves and needs an
electrical current, this electric current is weaker than the one in household wiring.
In EMG, the electrical instincts are picked up by needle conductors inserted into
the muscles and augmented on an oscilloscope display in the form of wavelike
tracing. The visual recording may go along with acoustic monitoring in which
sounds are augmented. It converts the minute muscular activities that occur when
we speak and converts them into electrical pulsations that can be transformed into
speech, without a sound coming out from mouth.
Fig: -Electromyographic Instruments Attached to Face [2]

4. Silent sound technology
Department of MCA 2015-16 Page 4
Fig: -Electromyography activity [3]
5.2 Image processing
Image processing is a process to transform an image into digital form and execute
some tasks on it, in order to get an enriched image or to extract some useful data
from it. It is a type of signal processing in which the input is image, like video
frame or snapshot and output may be image or features related with that image. In
silent sound technology the output will be in the form of audio. Visually image
dispensation system includes considering images as two dimensional signals.
Examination of remotely sensed data is done by means of different image
processing systems and methods. It includes:
5.2.1 Analog processing of image
5.2.2 Digital processing of image
5.2.1 Analog image processing
Analog image processing methods are applied to hard copy data, like printouts and
pictures. Image experts use various fundamentals of understanding while using
these visual techniques. It adopts certain rudiments of interpretation, such as
primary elements, three-dimensional arrangement etc.
With the grouping of multi-perception of inspecting remotely sensed data in
multispectral, multi-progressive, multi-scales, and in combination with

5. Silent sound technology
Department of MCA 2015-16 Page 5
multidisciplinary, permits us to make a judgment not only as to what an entity is
but also its significance. Apart from these it furthermore comprises optical
photogrammetric systems allowing for accurate dimension of the altitude breadth
position etc. of an entity.
Fig: - Analog Signal Process [4]
5.2.2 Digital image processing
Digital image processing includes an assembly of techniques for the operations on
digital images by computers. It contains some imperfections. To overcome the
imperfections and paucities in order to acquire the originality of the data, it is
essential to go through numerous steps of dispensation.
Digital image processing experiences three general steps:
a. Pre-processing
b. Enhancement and display
c. Information extraction
Fig: - digital signal process [5]

6. Silent sound technology
Department of MCA 2015-16 Page 6
6. ARCHITECTURE
The process model assumed is iterative process model since it is more adaptable
for this work. Once the face detection and the mouth region detection is achieved,
speech analysis can be performed with the use of lip motion features strategies and
emotional expressions with the use of other facial parts. If efficiency with
identification technique is not proper then the threshold value falls out of the
defined unique index value and retrial has to be made.
As the live video is captured by a high resolution camera, the video can be
processed as normal or grey scale color mode for customization. Region of interest
(ROI) video is segmented from which facial features like mouth, nose and eyes are
detected.
Fig: - Process Model Architecture And Its Working Methodology [6]
As the lip contour is initiated accurately, the extracted lip contour is
morphologically processed and corners are fitted by key points. A multi-frame
montage in a single object image montage is built and a database is created for it
and obtained features like eyes and nose vector in the database. The unknown

7. Silent sound technology
Department of MCA 2015-16 Page 7
templates of user are then compared with the existing template in database. If the
frames pass the threshold with the known and unknown templates, based on tested
index the user can receive a text output and an audio output.
7. APPLICATIONS
The technology has numerous applications such as stated below:
Assisting persons who have lost their vocal sound due to illness or misfortune and
person having speaking disorder so they can use this silent sound technique to
overcome their disability of aphasia. Silent sound technique is applied in military
for communicating underground/private matter which is very sensitive if leaked
can cause drastically damage to national security. We can make soundless calls
even if we are positioned in a jam-packed place such as cinema, trains, bus etc.
Useful for astronauts since sound needs a medium to travel and there is no medium
in space so by using silent sound technology astronauts can communicate with
each other in the space easily.
8. FUTURE PROSPECTS
Silent sound technology leads way to a shining future yet to come to speech
recognition technology. Without having conductors hanging all around your face
these conductors will be combined into cell phones. It might have features like lip
reading grounded on image recognition and dispensation rather than
electromyography. Also decreasing the time it took to process the photograph in
digital image dispensation. In future this technology will be of daily use. Engineers
claim this technology may be of daily use after five or 10 years. The technology
will have great impact on the mankind in coming years.
9. ADVANTAGES
It is very useful for those persons who have lost their voice and has been rendered
mute due to accident.
At public crowded places like in market, bus, train, malls, theatre etc. this
technology will be useful to talk without having surrounding noise problems.

8. Silent sound technology
Department of MCA 2015-16 Page 8
Very good technology for noise cancellation technique. Helps in making phone
calls in nasty environment.
Very useful for sharing confidential information like secret PIN number on phone
at public places.
It is very useful for astronauts to communicate with each other in space as there is
no medium for the sound to travel so this technique will come handy there.
10. DISADVANTAGES
This device at present needs nine leads to be invoked to our face which is quite
impracticable to make it usable.
Complicated to identify person and emotions. Technology is difficult to apply in
some languages like Chinese where dissimilar tones can hold unlike meanings.
From security point of view recognizing whom you are talking to gets complicated.
Every time you will feel like you are talking to a robot.
11. CONCLUSION
Silent Sound Technology is one of the modern developments in the field of
information technology, implements “Talking without Talking”. Engineers claim
that the device is functioning with 99 percent effectiveness. It will be one of the
ground-breaking and beneficial technologies and in forthcoming future this
equipment will be part and parcel of everybody’s life.

9. Silent sound technology
Department of MCA 2015-16 Page 9
12. REFERENCES
[1] http://www.scribd.com/doc/70178742/silent-sound-technology#scribd
[2] http://www.authorstream.com/presentation/DMmemon-1714484-silent-
Sound-technology
[3] http://www.telecomspace.com/content/cebit-2010-silent-soundtechnology-
Endless-possibilities
[4] http://www.techpark.net/2010/03/04/silent-silent-technology-an-end-to-
Noisy-communication
[5] http://www.dellchallenge.org/projects/silent-sound-technology
[6] http://www.whytelecom.com/content/cebit-2010-silent-sound-technology-
Endless-possibilities
[7] Shehjar Safaya , Kameshwar Sharma , Silent Sound Technology- An
End to Noisy Communication, Speech Communication , Vol. 1, Issue 9,
November 2013
[8] Denby B, Schultz T, Honda K, Hueber T, Gilbert J.M., Brumberg J.S.(2010).
Silent speech interfaces. Speech Communication
[9] Evangelos Skodras and Nikolaos Fakotakis, An Unconstrained method for lip
detection in color images by, IEEE ICASSP, ISSN : 1520-6149,pp: 1013 –
1016, 2011.
[10] Jian-Gang Wang, Eric Sung, Frontal-view face detection and facial feature
extraction using color and morphological operations, Journal Pattern
Recognition Letters archive, Volume 20 Issue 10, Oct. 1999, Pages 1053