Brain Computer Interaction of Indian Facial Expressions Recognition Through Digital Electroencephalography by ijcsis


More Info
									                                                                  (IJCSIS) International Journal of Computer Science and Information Security,
                                                                                                                  Vol. 9, No. 2, February 2011

  Brain Computer Interaction Of Indian Facial Expressions Recognition Through
                        Digital Electroencephalography

                  Mr.Dinesh Chandra Jain                                                                Dr. V.P Pawar
  Univ. of RGPV, Dept. Of Computer-Sc & Engineering                                     Univ. of Pune, Dept. of Computer App.
            Shri Vaishnav Inst. of Technology                                           Director of Siddhant Inst. of Comp-App
                      Indore, India                                                                    Pune, India

Abstract— The brain computer interaction could be the interface                so expression may consider as a vector in the seven
medium of the future, instead of using peripheral input output devices         dimensional field[5]. In anatomical FACS , the subject have
.So The brain computer interaction is a path way in which through              been made to express expressions which will have anatomical
digital EEG technique the brain signals of human subject have been             aspects of the face. The facial expression are controlled by the
recorded     under      different     poses    by    using     Digital
                                                                               brain so it is useful to correlate the expression with the
Electroencephalograph (EEG) 2400NP instrument. Under
experimental setup The subjects have given different expressions
                                                                               brain[6]. Today’s there are so many techniques available for
corresponding brain signals that have been recorded through a                  direct contact with neural as Electroencephalography(EEG),
popular technique Digital EEG. An attempt has been done to                     Magneto encephalography (MEG) and FMRI[7].
correlate these results to the facial action coding System (FACS).
                                                                                                  II.     BCI THROUGH EEG
                                                                                   In brain computer interaction electroencephalography
   Keywords- Bci, Eeg, Expression, Facial coding System.                           technique     is an approximation of the cumulative
    All standard paper components have been specified for                          electrical activity of neurons and is a measure of the
three reasons: (1) ease of use when formatting individual                          brain's voltage fluctuations as detected from scalp
papers, (2) automatic compliance to electronic requirements                        electrodes. This technology is to augment human
that facilitate the concurrent or later production of electronic                   capabilities by enabling human subject to interact with a
products, and (3) conformity of style throughout a conference                      computer through a conscious and spontaneous
proceedings. Margins, column widths, line spacing, and type                        modulation of their brain waves after a short training
styles are built-in; examples of the type styles are provided                      period. A brain computer interaction has been developed
throughout this document and are identified in italic type,                        cerebral electric activity is recorded via the
within parentheses, following the example. Some components,                        Electroencephalography: electrodes , attached on the scalp
such as multi-leveled equations, graphics, and tables are not                      and measure the electric signal of the brain[8] . The
prescribed, although the various table text styles are provided.                   signals are amplified and transmitted to the computer
The formatter will need to create these components,                                which transform them into device control command. The
incorporating the applicable criteria that follow.                                 crucial requirement for the successful functioning of the
                                                                                   brain computer interaction is that the electric activity in
                      I. INTRODUCTION                                              the scalp surface.
    The most important challenging application of        brain
computer interaction is to enables direct interaction between
human and computer by directly receiving and transmitting
signals to and from the brain of human subject. In the
Computer system the Recognition of Facial expression of
human subject is a great challenging task[1][2][3].
During expression recognition of human face there are so
many complex issue arises like: neuromedical ,anatomical
and psychological[4]. This is dependent on the social behavior
of human.. Basically, one human being may have different
expression under different -2 conditions. The human subject
may have seven type of universal facial expressions like:
happiness, sadness, fear, anger, surprise, disgust, and neutral                                 Fig1. Brain computer Interaction
   Identify applicable sponsor/s here. (sponsors)

                                                                                                            ISSN 1947-5500
                                                                        (IJCSIS) International Journal of Computer Science and Information Security,
                                                                                                                        Vol. 9, No. 2, February 2011

                      EEG frequency band

There are Five rhythms as:
1) Gamma ϒ: 30-50 Hz 2) Beta β: 13-30 Hz 3) Alpha α: 8-
13 Hz 4) Theta θ: 4-8 Hz 5) Delta δ: 0.5-4 Hz.

                      EEG Characteristics

 • It measures directly brain function.
 • It has a high temporal resolution, in the range of
 • The spatial resolution is in the range of centimeters for
   scalp electrodes, while implanted electrodes can measure                                   Fig 4:The international 10-20 system
   the activity of single neurons.
 • Scalp electrodes are non-invasive while implanted                              The test was conducted for ten minutes and each participant
   electrodes are invasive.                                                       was asked to give the different expressions while imaging the
 • The required equipment is portable.                                            particular    situations   portraying    different   emotions
                                                                                  simultaneously. The lower filter of the Neuro portable EEG
               Experimental method and procedure                                  was set at 1Hz, High filter at 70Hz; sensitivity at 7μV,
                                                                                  channels 20, sweep speed 30mm/sec, Montage set 1 for all
During eeg test the subject should be prepared to give                            experiment[13]. The resultant facial expressions of the
different type of facial expressions[9]. To conduct test , the                    participants were also captured photographically with the help
scalp must be free from oil now tie The electrode cap over                        of a digital camera. At the same time, the signals of the
head with the use of electrode cream and finally check all the                    different regions of the brain were mapped. The different
electrodes are connected with subject head, now said to                           position of connection for signal monitoring is shown in fig
subject to give six universal face expression[10] and on                          4[14]. An example of mapped signal is shown in fig 5. The
behalf of expression see the signal fluctuation and record it                     experiment was conducted in the sound proof environment.
with the four different regions of the brain[11].

    Fig2: Electrode connected on head.   Fig 3 . Four region of brain

                                                                                          Fig5 : signal of different -different brain area.
   In our experimental setup we have to selected fifteen male
persons in the age group of 6 to 30 year with non-psychiatric
history have been selected for our experiment. The electrode
cap was placed on the different regions of each person EEG
was recorded at sites of brain region for all frontal lobe and
parietal lobe are put together FP2-F4(R), FP1-F3 (L) and
Occipital lobe (C4-P4(R), C3-P3 (L)/Temporal lobe are kept
separate. FP1, FP2, F3, F4, F7, F8, FTC1, FTC2, C3, C4, T3,
T4, TCP1, TCP2, T5, T6, P3, P4, PZ, O1, O2,A1,A2. The fig
4 contains details of connection for left and right portion of the
brain in 10-20 international system[12].

                                                                                              Fig 6: Eeg signal window during test

                                                                                                             ISSN 1947-5500
                                                                 (IJCSIS) International Journal of Computer Science and Information Security,
                                                                                                                 Vol. 9, No. 2, February 2011
The signals corresponding to different expressions were                     The description of the apparatus used as Electrode cable, GND
recorded and stored in the computer[15] .The four regions of                plug wire, Phonetic wire, EEG conducting paste, Absorbent
the brain with left, right, front and back positions are identified         cotton wool I.P., Sony digital camera HD 5X.
corresponding to these positions, the average frequency and
the average peak voltage are determined through commercial                                   III. Result and Conclusion
software available with the system The signals were recorded
three times for each expression corresponding to all different              The goal of the present research work is to represent the
seven subjects. The average values are mentioned in the Table               experimental work towards bci direction. In our experimental
1.The EEG 24/NP channel unit used was from Digital                          setup, A portable Electroencephalograph system has been used
Neurocompact Medicaid System[16].                                           for brain. It has been found that in the human subject there are
                                                                            twenty six muscles are responsible for the overall movement
                                                                            of the face. The photographic expressions for each experiment
                                                                            are shown in fig (7).

                     Happy            Sad                 Angry                  Fear              Disgust               Surprise

                                                           Fig 7. Six Facial Expression

                                                                                                                         All Four
                 Brain                                                                                                  Positions
                Regions                     Position 1                               Position 2                           Result
                                     FL                 PL                    OL                       TL              +PL+OL+TL
                                     Left              Right                  Left                    Right            Left & Right
                                                                                                                       Avg. Avg.
                                                                                                                        of      of
              Expressions        F      PV         F       PV         F           PV              F           PV        F      PV
                                24.5 38.6        22.5     34.0     21.5          34.5           20.5          37       23.0       35.15
                                 24 44.5         21.9     42.1     24.5          36.2           23.1          32.9     23.37 42.75
                 3.Sad          22.5 34.3        26.3     39.6     20.2          33.0           22.1          36.8     25.35 35.92
                                19.6 35          26.1     35.6     23.4          31.1           20.4          33.4     22.37      33.77
                 5.Fear         19.7 36.8        18.9     34.9     16.6          27.0           16.9          36.1     19.11      35.37
                6.Disgust       19.5 40.2        20.1     35.2     18.5          30.2           19.6          38.2     19.42      35.95
               7.Surprise       13.1 32.4        17.9     34.7     13.9          32.8           13.8          34.7     14.67      33.65

                                                 Table 1. Results for different facial expression

                                                                                                        ISSN 1947-5500
                                                            (IJCSIS) International Journal of Computer Science and Information Security,
                                                                                                            Vol. 9, No. 2, February 2011


  [1]. Desney S. Tan, Anton Nijholt(Eds.)” Brain computer             [16]. Clinical application of an EEG-based brain-computer
 interaction applying our minds to HCI”,springer                      interface: a case study
  [2]. Rosenberger E. and Ekman, E.1994, Coherence between            in a patient with severe motor impairment by Neuper C,
     expressive and experimental system in emotion                    Müller G, Kübler A, Birbaumer N, Pfurtscheller G. Clinical
    cognition. Emotion, 8, 201-229.                                   Neurophysiology, 114(3):399-409 (2003)
  [3]. Cohn, J. & Zlochower , A.J. & Lien, J.J. & Kanade
   T.(1998). Feature- by optimal flow discriminates subtle            [17]. D. Gutiérrez, F. García-Nocetti, and J. Solano-Gonzalez,
   differences in facial expression. Third IEEE International         “Classifififification of Multichannel EEG Data
  conference on Automatic Face and Gesture Recognition.               using Lenght/Energy Transforms,” in Proceedings of the 2005
    396-401.                                                          1st IEEE International Workshop on
 [4].J.DBayliss,”A Flexible Brain Computer Interaction “,             Computational Advances in Multi-Sensor Adaptive
    university of Rochester,2001.                                     Processing, 2005, pp. 221–224.
[5]. Wheeler, R.E., R.J. Davidson and Tomarken , A.J.1993,
   Frontal brain asymmetry and emotional reactivity:                                          Author Profile
 biology substrate of affective style, psychophysiology ,82-89.
[6]. P.Ekman,”Recognition of six basic facial expression and
 their strength by neural “International workshop in
       IEEE 1992.                                                                      Mr. Dinesh Chandra Jain has completed B.E
  [7]. Cohn, J.A., Allen J.J.B and Harmon- Jones, 2001,                                (Comp-Sc) in 2004 and M.Tech (IT) degree in
                                                                                       2007. He is presently               working as
     Voluntary facial Expression and hemispheric asymmetry                             Assistant Professor in the Department of Comp.
       over the frontal cortex, Psychoysiology, 38, 912-925 .                          Science & Engg. at SVITS, Indore. and He is
  [8] .Terzo poulous, D. & Waters, K. (1993). Analysis and                             pursuing PhD in the field of Digital Image
 synthesis of facial image sequences using physical and                                Processing       &         Neural      Network
       Anatomical models. IEEE Trans.Pattern analysisNetwork.
                                                     Neural   and
 machine intelligence 15, 6: 569-579.
 [9] Blundell, G. G. The meaning of EEG. London:
Audio Ltd.                                                                        Dr.Vrushsen Pawar received MS,
 [10]. Frances M.Dyro,”The EEG Handbook” clinical                                 Ph.D.(Computer) Degree from Dept .CS
neurophysiology,LaboratoryMassachusetts, London,1989.                             & IT, Dr.B.A.M. University & PDF from
   transactions on neural system.                                                 ES, University of Cambridge, UK. Also
[11].Ying-li Tian,Takeo Kanade J.F. Cohn “Recognizing
                                                                               Received MCA (SMU), MBA (VMU)
Action Unit for facial expression analysis”, IEEE                                 degrees respectively. He has received
                                                            prestigious fellowship from DST, UGRF (UGC), Sakaal
      transactions on Pattern analysis & machine intelligence,
Feb(vol-23,no.2) pp. 97-115,2001                            London, ABC (USA) etc. He has published 90 and more
                                                  foundation, ES
[12].Delorme and S.Makieig,”EEg changes accompanying        research papers in reputed national international Journals &
learned regulation of 12-43 EEG – activity “ IEEE           conferences. He has recognize Ph.D Guide from University of
[13]. The "10-20 System" of Electrode Placement             University & Sighaniya University (India). He is senior IEEE
                                                    une, SRTM
                                                             member and other reputed society member. Currently working             as a Professor & Director in SICA institute is affiliated to
                                            University of P pune.
[14]. Gupta, S.; Singh, H. Preprocessing EEG signals for
direct human-system interface.
Intelligence and Systems, 1996., IEEE International Joint
Symposia on , 1996
Page(s): 32 –37.

[15]. T.M. Vaughan, J.R. Wolpaw, and E. Donchin, "EEG-
Based Communication:
Prospects and Problems", IEEE Trans. on Rehabilitation
Engineering,4:4,pp. 425—430,1996.

                                                                                                 ISSN 1947-5500

To top