Jan 28, 2015
Silent Sound Technology
Najeeb P [EPAKECS029]
Guide : Raseek C , Asst.ProfessorDepartment of Computer Science and Engineering
Government Engineering College,Sreekrishnapuram
12th March 2014
Introduction
Why ..... ???
Methods
Design
Features
Restrictions
Applications
Conclusion
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 2
• Silence is the best answer for all the situations... even yourmobile understands!
• A technolgy that transforms the information without using ourvocal chords.
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 3
Introduction
Why ..... ???
Methods
Design
Features
Restrictions
Applications
Conclusion
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 4
Need For Silent Sound Technology
1
1http://www.go-gulf.com/crowd talking on mobile phones-300x225.jpgSilent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 5
Need For Silent Sound Technology
• For a better way of communication in noisy public place.
• Perfect solution for those who lost their voice.
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 6
Introduction
Why ..... ???
Methods
Design
Features
Restrictions
Applications
Conclusion
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 7
Two methods are used
• Electromyography(EMG)
• Image Processing
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 8
Electromyography(EMG)
2
2http://www.telecomspace.com/�les/cebit/device.jpgSilent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 9
Electromyography(EMG)
• Even by saying words without producing sound weak electriccurrents are sent from brain to speech muscles. Thisphenomenon is called ELECTROMYOGRAPHS.
• A technique for evaluating and recording the electrical activityproduced by articulatory muscles.
• The transducers involved convert the pulses into electricalsignals.
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 10
Electromyography(EMG)
3
3Reference[1]Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 11
Working..
• The sensors(each with two �ne-wired electrodes) are attachedto the di�erent part of the face
• The electrical activity is recorded
• The sensors has to be placed at various locations to obtain anaccurate signal
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 12
Working...
• Vocoder (voice encoder) converts this electrical signals intospeech patterns.
• Compare them with pre recorded speech pattern of spokenwords
• When there is a match, that sound is transmitted on to theother end of the line.
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 13
Sensors used
• Pressure sensors
• Vibration sensors
• Electromagnetic sensor
• Motion sensors
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 14
Image Processing
4
4http://www.telecomspace.com/silent sound/image processing.pngSilent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 15
Image Processing
• Converting digital data tape into a �lm image with minimalcorrections and calibrations.
• In the silent sound technology the output of this imageprocessing is an audio record.
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 16
Introduction
Why ..... ???
Methods
Design
Features
Restrictions
Applications
Conclusion
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 17
Components
• Data acquisition
• Online recognition
• Sound playback or synthesis
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 18
EMA based silent speech interface
5
5Reference [1]Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 19
Data Collection and Processing[1]
Stimuli and Procedure
• A �xed number of sentences for basic greeting andconversation were selected form a list of most frequently usedsentences among AAC users
• The electromagnetic articulograph AG500 was used to registerthe 3-D movements of the tongue, lip, and jaw when a subjectwas talking.
• The articulograph records the movement by establishing acalibrated electromagnetic �eld.
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 20
Data Collection and Processing[1]
Data Processing
• Head movements were subtracted from the tongue and liplocations.
• A zero-phase low-pass �lter was applied for removing signalnoise.
• All sequences were manually segmented and annotated withsentences.
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 21
Introduction
Why ..... ???
Methods
Design
Features
Restrictions
Applications
Conclusion
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 22
Features
• Native speakers can silently utter a sentence in their language,and the receivers can hear the translated sentence in theirlanguage.
• It appears as if the native speaker produced speech in a foreignlanguage.
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 23
Features
• The translation technology works for languages like English,French and German.
• Allow people to make silent calls without bothering others.
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 24
Introduction
Why ..... ???
Methods
Design
Features
Restrictions
Applications
Conclusion
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 25
Restrictions
• This device presently needs nine leads to be attached to ourface which is quite impractical to make it usable.
• Complicated to identify person and emotions.
• Technology is di�cult to apply in some language like Chinese,where di�erent tones can hold many di�erent meanings.
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 26
Introduction
Why ..... ???
Methods
Design
Features
Restrictions
Applications
Conclusion
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 27
Applications
• Helping people who have lost their voice due to illness oraccident.
• Telling a trusted friend your PIN number over the phonewithout anyone eavesdropping � assuming no lip-readers arearound.
• Silent Sound Techniques is applied in Military forcommunicating secret/con�dential matters to others.
• E�ective calls in crowded place.
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 28
Introduction
Why ..... ???
Methods
Design
Features
Restrictions
Applications
Conclusion
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 29
Conclusion
• Silent Sound Technology one of the recent trends in the �eld ofinformation technology implements �Talking without Talking�.
• It will be one of the innovation and useful technology and incoming future this technology will be use in our day to day life
• The online algorithm still has room for improvement.
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 30
REFERENCE
• [1] Ashok samal,Frank Rudzicz,Jordan R Green and Jun Wang,�Sentence recognition from articulatory movements for SilentSpeech Interfaces� ,IEEE International Conference onAcoustics, Speech, and Signal Processing (ICASSP), 2012 ,Page(s): 4985 - 4988 .
• [2] Ashok samal,Jordan R Green, Jun Wang and Tom D.Carrell, �Vowel recognition from continuous articulatorymovements for speaker dependent application�, SignalProcessing and Communication Systems (ICSPCS), 2010 ,Page(s): 1-7.
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 31
• [3] Guangqi Cao, Meng M. Helen and Zhiyong Wu,�A Uni�edFramework for Multilingual Text-to-Speech Synthesis withSSML Speci�cation as Interface�,IEEE Tsinghua Science andTechnology ,2009 , Page(s): 623 - 630 .
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 32
Thank You for the attention
Silent Sound Technology Najeeb P [EPAKECS029] Department of Computer Science 33