lifeasIknow-IT

PAM Face Authentication Musings: How To Use

2009-08-09T17:10:00.000-07:00

http://code.google.com/p/pam-face-authentication/

Important: Update 30th August -

* [ changed CMAKE_INSTALL_DIR to CMAKE_INSTALL_PREFIX ]

I have got it working with Pardus 2009 , for other distros , some of the configuration that is given below might change.

Pardus

http://distrowatch.com/table.php?distribution=pardus

What you Need

for pam-face-authentication-0.3.tar.gz

Opencv 1.0.0 +

libpam ( also the devel package)

libX11 (also the devel package )

Qt4.5 (Not tested with lower versions)

for kdm-plugin-pam-face-authentication-0.1.tar.gz

libX11 (also the devel package )
KDE4 (and dev package )

Install From Source

wget http://pam-face-authentication.googlecode.com/files/pam-face-authentication-0.3.tar.gz

tar -xf pam-face-authentication-0.3.tar.gz

cd pam-face-authentication-0.3

mkdir build

cd build

cmake -D CMAKE_INSTALL_PREFIX=/usr ..

make

sudo make install

wget http://pam-face-authentication.googlecode.com/files/kdm-plugin-pam-face-authentication-0.1.tar.gz

tar -xf kdm-plugin-pam-face-authentication-0.1.tar.gz

cd kdm-plugin-pam-face-authentication-0.1

mkdir build

cd build

cmake -D CMAKE_INSTALL_PREFIX=/usr/kde/4 ..

make

sudo make install

Install From Pardus Contrib Repo

I will update this place when the package is tested etc.

How to Train your face

After Installation , you can run qt-facetrainer,

a desktop file is added during installation, so you can search for "Face Trainer" under applications , or run qt-facetrainer from Konsole

How To Configure

for Face Authentication with - su

Edit /etc/pam.d/su and add the following line at the top

auth sufficient pam_face_authentication.so

instead you can add the following line to get Webcam Preview Window

auth sufficient pam_face_authentication.so enableX

Configuring KDM to use KDM Greeter Plugin

Edit /usr/kde/4/share/config/kdm/kdmrc

search for PluginsLogin

uncomment the line and add

PluginsLogin=classic, faceauthenticate

NOTE: It seems the plugin doesnt work in some Themes ( like air from pardus)
so i am using oxygen, it works on circles also [ Current task is identify why its not working on all themes ]

Theme=/usr/kde/4/share/apps/kdm/themes/oxygen

If you are sitting at locations where there is not enough illumination on the face, the monitors light actually makes a difference , You should change the KDM theme to some very light ( white ) background which will illuminate your face.

Create a new file /etc/pam.d/kde4-faceauthenticate with the following contents

#%PAM-1.0

auth sufficient pam_face_authentication.so

account include system-auth

password include system-auth

session include system-auth

session optional pam_console.so

session optional pam_polkit_console.so

WARNING:this configuration file require in Pardus 2009 , this could change, it could be kdm-faceauthenticate in other distros ( Will Update after checking other distros )

Configuring GDM (older GDM versions <2.24)

Edit /etc/pam.d/gdm and add the following line at the top

auth sufficient pam_face_authentication.so enableX

(enableX options works in Ubuntu 8.04 and Debian Sid, I have not tested with other distros, if this doesnt work , you could try changing enableX to gdmlegacy )

Configuring GDM (newer GDM versions > 2.24 )

I really did not have time to do this ( maybe it works ) [ There is a hack to get it working ill update this space after I test it ]

Important:

Bug #1,

Currently with KDM, if you choose face authentication plugin and fail to login, and if you try switching back to Classic Password Plugin , the password box doesnt accept anything from the keyboard

if you are not able to login with face authentication plugin , you can kill X (pkill X from terminal)

it will restart the KDM with classic plugin

Bug #2,

Screen Lock in KDE needs a patch to get it working with the plugin ( Oswald Buddenhagen – The KDM Maintainer gave a patch , but somehow I couldn't get it working , will need to ask him again when he is free.)

Bug #3, Screen Lock in GNOME, does work, but doesnt show the Webcam Preview Window ( I think there is a patch somewhere , that fixes this ( will update later))

HAVE FUN!

PAM Face Authentication Musings: The UI and all things nice

2009-08-05T17:40:00.000-07:00

This is much awaited post, i am back to college and having the best semester ever till now =), No more nonsense ,straight to the point ... here are the screenshots

People to Thank For Getting the Project Till this point=)
Onur Küçük - My Soc Mentor under Pardus =) (THANKS )
Pinheiro - Giving Advice to remove custom stylesheets (that makes sense)
Alex Lau - My Soc 2k8 Mentor ( THANKS )
Nikolay Shmyrev - ( without him the project would have seen the light of day last year, THANKS )
Ossi - The KDM Maintainer ( I bugged him too many times :P THANKS)
Ashish Gandhi - Hindi Translation ( धन्यवाद :P )

The First Screen ... ( Read'em Carefully)

whoooops Webcam Not connected!

after connecting the webcam

Train Your Face - Create Sets in Different Illumination

Advanced Config - Tryout recognition , Decrease Security Level , Set Defaults

wooooo, you are done , off you go to your KDM , GDM or ....

PS: Docs are here
http://www.lifeasiknow-it.com/projects/pam-face-authentication/docs/

A Very Sexy Collaboration Diagram is here
http://www.lifeasiknow-it.com/projects/pam-face-authentication/docs/html/classfaceTrainer.html

Coming Up: KDM Video Preview, Instructions on How to Use, Packaging

PAM Face Authentication Musings: Face Verification using cascaded MACE and LBP based Classifier

2009-07-28T18:22:00.000-07:00

http://code.google.com/p/pam-face-authentication/

By Rohan Anil, BITS Pilani Goa Campus

Our objective is to verify whether the given test image of face belongs to a particular class ( a user ) , in other words to check if it is what it claims to be.I will be introducing two methods and how I cascaded it to produce high Verification rate. The Cascading helped in reducing false acceptance ( an imposter getting verified , although the claim was false) and reducing false rejection ( a true user getting rejected as an imposter ).

MACE FILTER

Mace stands for Minimum Average Correlation Energy, it was introduced by Mahalanobis et al in 1987 ( way before I was even born ) , Its was applied to Face Verification by Vijayakumar , Savvides et all. They had successfully applied to CMU PIE Database Giving phenomenal results. Correlation can be described as matching , it has been used widely in Number Plate recognition et. It is a very easy concept, you can read it from almost every Introductory Image processing text books. One interesting thing about correlation is its equivalent in frequency domain is a simple element by element multiplication (taking the conjugate transpose of the first matrix).

The filter is produced through the following equation

There X is the Matrix d X N where N is the number of training images and d = image_width X image_height , and matrix D is a diagonal matrix of size d x d , the values are the average Fourier spectrum of training images. You can find different types of modification of the D matrix in the literature , one such is MINACE.

The h is a filter in frequency domain , now given a test image , for verification , we correlation, so in frequency domain, it becomes a simple multiplication.

Equation used for Correlation is iFFT( Conj_Transpose(FFT(TestImage)) .(X) h ) , this equation would give us the correlation output in Spacial domain. Now the next question is How do we verify? To do that we need a measure which shows how close the image is to the face of the claimed user. In the code there are two measures, one is Peak to side lobe ratio (PSLR) another is Peak correlation energy ratio (PCER), but currently only PSLR is used, since it gave higher verification rate on a small dataset than PCER. You can read the following paper [1] for the equation. Higher the PSLR the closer the image is to the face of the claimed user. The thresholds for PSLR is automatically (or automagically ) found from the training set itself. Problems I think with MACE filter is that, its a ratt-learner (ratt in hindi means by-heart ) , it just ratt's the the training images, which sometimes allows imposters to get verified. MACE filter also has problems coping with illumination datasets, and registration errors (eyes detected with >2% error (length) which seems small) can cause a very serious problem for verification. So for each illumination set a different MACE filter is trained.

LBP METHOD

LBP is a spatial filter(sort-of), it expands to Local Binary Pattern was introduced by Ojala et al. and can be used to get the texture features out of the Image. It is illumination tolerant to a great extent. It is very simple to compute. The operator assigns a label to every pixel of an image by thresholding the 3×3 neighborhood of each pixel with the center pixel value and considering the result as a binary number.A Local Binary pattern is uniform if the binary pattern contains at most two bitwise transitions from 0 to 1 or vice versa, in a circular fashion. This paper [2] would be an ideal way to start learning on its use.The following Images , represents the LBP output.

I have taken the average face of each Set and applied this operator and used a Chi Square distance (refer [2]) , if the test image falls below a certain threshold ( which was found through some analysis on a dataset) , this stage is declared as 'passed'.

CASCADING

A Picture Speaks a thousand words, I hope to make up for around 10-20 sentence that I have not written with the following picture.

The following method is giving less than 1% False Acceptance on a private dataset , i will provide the results on YALE dataset later ( as soon i get a new dell laptop , this one is dying )

[1] http://www.ece.cmu.edu/~kumar/Biometrics_AutoID.pdf

[2] http://www.idiap.ch/~marcel/professional/publications/marcel-ijivp-2007.pdf

PAM Face Authentication Musings: Eye Tracking using - Grid Search

2009-07-27T06:25:00.001-07:00

http://code.google.com/p/pam-face-authentication/

By Rohan Anil , BITS Pilani Goa Campus

Initially Haar Face detection is run on the Image, and the Haar Eye Detection is run on the out put of face detection, The theory on haar detection is very interesting, I will give you a brief idea , the method was devised by Paul Viola and Michael Jones. In the Initial Stage, thousands of false images and true images are taken. True image means the face images and false images mean background, this could be anything. You can find a lot of papers about how to generate these background images. Now Haar Features are calculated of the true and false Images, Features mean in layman terms, some set of real values that can describe the “Object” in question, We are dealing with face here. Now after these features are calculated , we have to find, which of these Features are good , assign some kind of weights to it. These weights are calculated using adaboost, developed by Freund and Schapire, its a very simple iterative algorithm, which tries to find the weights for the features by trying to minimize the number of misclassification at every stage. Finally, when you perform detection, a search is done on the image, for every window in the search, features are calculated , and the classifier , usually a distance formula output, is calculated and depending on whether its less than a particular threshold , its marked as a face or not face.A Similar process is run for eye detection on the detected face. For the project, I have selected only the biggest face that is detected on the image , since its an authentication project. After the eye is detected. A center of Mass of the Eye Window is Calculated by using (255-Intensity) as the mass. You will be surprised how this concept from Phy 101 , leads to good Localization.

Tracker Model initialization is done after this step, Integral and Variance Projection is calculated in X and Y directions. Integral projection is the sum of Pixels , and Variance projection is the variance of the pixel values. Please not that , the Intensity values are calculated as (255-Intensity), but the figures in the articles are shown with the actual Intensity. Now from the previous step the Anchor Point is taken which is the eye co-ordinates localized and given as a parameter in the initialization step. This process is repeated until the eye detection step fails, at the moment the tracking algorithm is used to localize the eye co-ordinates in the the frame. Before I explain how its done, I will have to introduce some concepts from Math 101, Matching two 2D functions,

y=F(x) is your equation , ie is say the Integral Projection of the model [1]

y'=G(x) is the equation for the Integral Projection in the new frame [2]

now we use two parameters the, the translation parameter and scale parameter, and we will define the Distance Formulae with [1] and [2]

distance = pow((ScaleFactor*G(x) + TranslateFactor – F(x)),2)

So our objective is to find the Scale Factor and Translate Factor which gives you the minimum distance and hence achieving good localization

So a Iterative Grid Search for Scale between (a1,a2) and Translation between (t1,t2) is used. Iterative Algorithm looks at all the points on the Grid, and then decreases the step size by Half , and continues the same process for K number of Iteration, where intial step size (a1-a2)/N and (t1-t2)/N where N =5 for a 5X5 grid size. The code currently uses scale between (.9,1.1) and translate factor between (-4,4) . Although I have not completely explained every part of the algorithm, that is it uses previous angle , width , height, and some other parameters making it a bit more complex to explain.

Here is the Youtube Video.

I am able to get around 10fps on my laptop which is 32bit dual core 1.7ghz from pentium, Currently I am exploring a different algorithm , which would give >30 Fps, but I haven't released any code yet. The above algorithm can be improved a little by using Integral Image from the face detection for eye detection and tracker initialization , OpenCV 1.1.0 supports CV_BIGGEST_OBJECT, that could be used in face detection ,currently because OpenCV 1.0 is more widely used and packaged, I preferred not to swtich.

If you liked the article dont forget to comment and digg it!

Pam Face Authentication Musings : Autotools and Qt Gui Project

2009-06-19T03:27:00.000-07:00

Qt projects are usually packaged using a nice utility called qmake.But i am more comfortable with the autotools way. So after some intensive google searching. I finally figure out how to add the UIC (User Interface (.ui) to C Code) and MOC (Meta Object Compiler - which handles the signals and slots)

Configure.ac

  

#                                               -*- Autoconf -*-

# Process this file with autoconf to produce a configure script.



AC_PREREQ(2.50)

AC_INIT(pam-face-authentication, 0.3,[http://code.google.com/p/pam_face_authentication/issues/list])

#AC_CONFIG_SRCDIR([pam_face_defines.h])

AM_INIT_AUTOMAKE([1.10.1])

 

AM_CONFIG_HEADER([config.h])





# Checks for programs.

AC_PROG_CC

AC_PROG_CXX

AM_PROG_CC_C_O

AC_PROG_LIBTOOL



# Checks for libraries.



AC_CHECK_HEADER(X11/Xlib.h, check_x11="yes",check_x11="no")

  if test x${check_x11} = xno ; then

 AC_MSG_ERROR([X11/Xlib.h Not Found.])



  fi

  if test x${check_x11} = xyes ; then

    AC_CHECK_LIB(X11, XOpenDisplay,

                 [ X11_LIBS="-lX11"

                   AC_DEFINE(USE_X11, 1, [Define to use X11 copy'n'paste]) ],

                 [])

 AC_SUBST(X11_LIBS)



  fi





AC_CHECK_LIB(pam, pam_start, ,AC_MSG_ERROR([PAM library missing]))

AC_CHECK_LIB([pam], [main],,AC_MSG_ERROR(libpam-devel >.99.8 Not Found))

AC_CHECK_HEADER([security/pam_modules.h], ,[AC_MSG_ERROR([PAM headers missing])])



PKG_CHECK_MODULES(QTGUI, QtGui)

AC_SUBST(QTGUI_CFLAGS)

AC_SUBST(QTGUI_LIBS)



PKG_CHECK_MODULES(QTCORE, QtCore)

AC_SUBST(QTCORE_CFLAGS)

AC_SUBST(QTCORE_LIBS)



PKG_CHECK_MODULES(OPENCV, opencv >= 1.0.0)

AC_SUBST(OPENCV_CFLAGS)

AC_SUBST(OPENCV_LIBS)



AC_CONFIG_FILES([

 Makefile

 ])

AC_OUTPUT

Makefile.am

  

DEBUG = -g -O0

INCLUDES = \

 -DVERSION=\"".3"\"  \

 -DPKGDATADIR=\""$(pkgdatadir)"\"  \

 -DBINDIR=\""$(bindir)"\"   \

 -DSYSCONFDIR=\""$(sysconfdir)"\"   \

 $(OPENCV_CFLAGS) \

 $(QTCORE_CFLAGS) \

 $(QTGUI_CFLAGS)



bin_PROGRAMS = qt-facemanager

qt_facemanager_LDFLAGS = -export-dynamic

qt_facemanager_SOURCES = main.cpp \

   qrc_graphics.cpp \

   opencvWebcam.cpp \

   detector.cpp \

   detector.h\

   faceDetector.cpp \

   faceDetector.h \

   eyesDetector.cpp \

   eyesDetector.h \

   tracker.cpp \

   tracker.h \

   utils.cpp \

   opencvWebcam.h \

   faceTrainer.h \

   faceTrainerAdvSettings.h \

   ui_faceTrainerAdvSettings.h \

   ui_faceTrainer.h



qt_facemanager_LDADD = \

   $(QTCORE_LIBS) \

   $(OPENCV_LIBS) \

   $(QTGUI_LIBS)



nodist_qt_facemanager_SOURCES = \

 moc_faceTrainer.cpp \

 moc_faceTrainerAdvSettings.cpp



#put this file in the package

EXTRA_DIST = \

 graphics.qrc





# Convert %.ui files into ui_%.h

ui_%.h: %.ui

 uic $< -o $@



# Create moc_%.cpp from *.h

moc_%.cpp: %.h

 moc $< -o $@



qrc_graphics.cpp : graphics.qrc

 rcc -name graphics graphics.qrc -o qrc_graphics.cpp





CLEANFILES =$(nodist_qt_facemanager_SOURCES) ui_faceTrainerAdvSettings.h ui_faceTrainer.h qrc_graphics.cpp

Orignal from http://qtnode.net/wiki?title=Qt_with_autotools

Opensource Accessibility

2009-05-22T14:34:00.000-07:00

Basically this post is for introducing two Projects to you , and maybe you can contribute to it!

First One MouseTrap created by Flavio

What is MouseTrap? (taken from webpage)

* In few words mouseTrap can be described as an a11y software developed to allow people with movements impairments access the computer. mouseTrap is written in python, based on the OpenCV library and uses image processing to translate the user's head movements into mouse events ( movements, clicks ) which allow users to interact with the different desktops managers and applications.

Some Videos are at http://youtube.com/profile_videos?user=flaper87

So go ahead contribute!

Second one is
VizAudio : this project is in its conceptual design stage!
person to contact suseROCKS at irc.freenode.net , he hangs around in #opensuse-project

What is VizAudio?
The main aim of this tool is to provide Visual Feedback in replacement of Audio feedback for desktop events, so that it could be easily identified. There is a potential market for such an enriched Desktop , and this tool would decrease potential users looking for costly alternative proprietary systems. Visual events would correspond to specific application sound event. Work on a unifying framework for Visual Feedback from sound layer, the Visual feedback system should be independent of (compiz/metacity etc) , so that it could be universally used, no matter what environment the desktop uses. Visual feedback should be configurable and pluggable so that project can get attention from other developers.Use of libcanberra etc which has to be decided before the Coding the period , through discussions.
Some effects which could be possible

Examples would be:

1. Screen dimming flicker
2. Hard screen flicker
3. Running lights around the border of the monitor
4. Graphic popup in designated area of screen
5. Really Transparent , Scaled Icon of the Application on the Center of the Screen , so you could feel which application gave the notification ( but all these should be configurable , user should be given a choice )

I will be also trying to contribute to these projects during summer!

A VERY AWESOME SUMMER AHEAD!

My first First(or any :P) Author Paper!

2009-05-22T14:28:00.000-07:00

My paper got accepted for oral presentation at
International Conference on Artificial Intelligence and Pattern Recognition 09 at Orlando,Florida !
it will be held on July 13-16th, 2009

Right now money $$$$ is the only problem ! Will have to figure stuff out!

and Yeah you are right! i am going NUTS!!!! yipeeee!!! wohooooooo!!!

Google Summer of Code 2009

2009-05-22T14:03:00.000-07:00

This post is a month late , was busy at college had my comprehensive examinations , farewell party and yeah my ankle was dislocated :P

I have been accepted for Google Summer of Code , under Pardus for continuing and integrating my Last years face authentication module. I will be mentored by Onur Küçük

ITS GOING TO BE AWESOME !

These are some of the Principles which will be mostly followed
1. Make It Configurable (for advanced user)
2. Donot Expect the user to know anything, so default configuration should work fine with just anyone
3. Write loosely coupled code
4. Write Documentation with the code!

Basically these are some points that i want to keep in mind while coding :) , so this will keep me in check

First Priority
Remove SVM from the CODE
Why?
It takes a lot of time to train, so a user gets irritated by the amount of time it takes!

... ok enuf for this post , ill put more details later

Update: SVM Classifier Results

2009-03-23T02:11:00.000-07:00

I have added the SVM Classifier (using Libsvm) to the existing pam-face-authentication code, its in the svn and NOT ready for a checkout , there are still some importants tasks to finish

#) Remove exisiting Unused functions
#) Remove the unmaintainable libfacedetect.c and Add the lucas kanade tracker module
#) Use Kmeans and Distance Measure to reject Imposter Faces ( this is just an idea ) dont know whether it will work

I will update this space very soon

DctMod2 Feature For Face Recognition

2009-03-16T12:37:00.000-07:00

There has been an lot emphasis on the following feature in face recognition papers after 2002
It seems the recognition rate increases whatever classifier you use with the feature
Following is the C Code To be used with Opencv

You can refer the following paper

Fast feature extraction method for robust face verification
C. Sanderson and K.K. Paliwal

http://maxwell.me.gu.edu.au/spl/publications/papers/el02_conrad.pdf

Another Paper by Sanderson and Bengio is there on DCTmod3 coefficients

But Today DCTMOD2 rules

  

#include "cv.h"

#include "cvaux.h"

#include 

#include 



void  featureDctMod2(IplImage * img,float*  features_final)

{

    int Nx = floor((img->width - 4)/4);

    int Ny= floor((img->height - 4)/4);

    float* observations = (float*)malloc(  Nx*Ny * 25 * sizeof(float) );

    float* features = (float*)malloc(  Nx*Ny * 18 * sizeof(float) );

    cvImgToObs_DCT(img,observations, cvSize(8,8),cvSize(5,5),cvSize(4,4) );

    int i,j,k=0;

    for (i=0;i    {

        for (j=0;j        {

/*

0   1   2    3    4

5   6   7    8    9

10  11  12   13  14

15  16  17   18  19

20  21  22   23  24



Zig Zag Pattern

0,1,5,10,6,2,3,7,11,15,20,16,12,8,4



*/



// Zig Zag 15 features + 3 for dct mod2

            features[i*18*Nx + j*18 +0]=observations[i*25*Nx + j*25 +0]; 

// not putting zero , might cause scale problem

            features[i*18*Nx + j*18 +1]=observations[i*25*Nx + j*25 +1];

            features[i*18*Nx + j*18 +2]=observations[i*25*Nx + j*25 +5];

            features[i*18*Nx + j*18 +3]=observations[i*25*Nx + j*25 +0];

            features[i*18*Nx + j*18 +4]=observations[i*25*Nx + j*25 +1];

            features[i*18*Nx + j*18 +5]=observations[i*25*Nx + j*25 +5];

            features[i*18*Nx + j*18 +6]=observations[i*25*Nx + j*25 +10];

            features[i*18*Nx + j*18 +7]=observations[i*25*Nx + j*25 +6];

            features[i*18*Nx + j*18 +8]=observations[i*25*Nx + j*25 +2];

            features[i*18*Nx + j*18 +9]=observations[i*25*Nx + j*25 +3];

            features[i*18*Nx + j*18 +10]=observations[i*25*Nx + j*25 +7];

            features[i*18*Nx + j*18 +11]=observations[i*25*Nx + j*25 +11];

            features[i*18*Nx + j*18 +12]=observations[i*25*Nx + j*25 +15];

            features[i*18*Nx + j*18 +13]=observations[i*25*Nx + j*25 +20];

            features[i*18*Nx + j*18 +14]=observations[i*25*Nx + j*25 +16];

            features[i*18*Nx + j*18 +15]=observations[i*25*Nx + j*25 +12];

            features[i*18*Nx + j*18 +16]=observations[i*25*Nx + j*25 +8];

            features[i*18*Nx + j*18 +17]=observations[i*25*Nx + j*25 +4];

            for (k=0;k<18;k++)

            {

                   features_final[i*18*Nx + j*18 +k]= features[i*18*Nx + j*18 +k];

            }

        }



    }

    for (i=1;i    {



        for (j=1;j        {

   //Calculate Horizontal Delta and Vertical Delta

            features_final[i*18*Nx + j*18 ]=(-features[i*18*Nx + (j-1)*18 +3] +features[i*18*Nx + (j+1)*18 +3])/2;

            features_final[i*18*Nx + j*18 +1]=(-features[i*18*Nx + (j-1)*18 +4] +features[i*18*Nx + (j+1)*18 +4])/2;

            features_final[i*18*Nx + j*18 +2]=(-features[i*18*Nx + (j-1)*18 +5] +features[i*18*Nx + (j+1)*18 +5])/2;

            features_final[i*18*Nx + j*18 +3]=(-features[(i-1)*18*Nx + j*18 +3] +features[(i+1)*18*Nx + j*18 +3])/2;

            features_final[i*18*Nx + j*18 +4]=(-features[(i-1)*18*Nx + j*18 +4] +features[(i+1)*18*Nx + j*18 +4])/2;

            features_final[i*18*Nx + j*18 +5]=(-features[(i-1)*18*Nx + j*18 +5] +features[(i+1)*18*Nx + j*18 +5])/2;

        }

    }

free(features);

free(observations);

}



void main()

{

IplImage * img = cvLoadImage("sample.jpg",0);



/* Maximum number of Blocks */

int Nx = floor((img->width - 4)/4);

int Ny= floor((img->height - 4)/4);

/* Feature first 6 coefficients are the delta coefficients and rest 12 are from the zig zag order */

float* features = (float*)malloc(  Nx*Ny * 18 * sizeof(float) );



featureDctMod2(img,features);

cvNamedWindow( "Source", 1 );

cvShowImage( "Source", img  );

cvWaitKey(0);

}

Ankle Dislocation and a Ligament(s) tear And Other Updates

2009-03-10T12:32:00.000-07:00

While playing basketball , it happened, (Ankle dislocation Inversion):( , need 4-5 weeks for the plaster to come off , another 3-4 months to heal , i am hearing rumours of ligaments not coming back to its 100 % , bah!!

in other news,

I joined Face recognition For Facebook community ! , i feel its going to be awesome contributing to that project,
I did apply to IAS Summer research fellowship , but did not get through(probably because i did not contact the professor)(i wanted to work on AAM's)
I am currently working on a fusion of
PCA + SVM
LDA +PCA +SVM
DCT+(PCA??) + SVM

classifier, i will keep updating the blog , with the results

and yes LK algorithm has been deciphered :P , check out my Interpreting Lucas Kanade Post for details( i will put it up soon)

Xauth Bug Squashed

2009-02-23T12:20:00.000-08:00

Xauth Bug has been squashed!!
Finally some progress, svn has been updated.

Bug Fix Method :-
getting Display using PAM function
checking if XopenDisplay is working
if not then
open /tmp/.X(dispNumber)-lock
get PID of xserver
extract xauth filename from /proc/PID/cmdline
setenv XAUTHORITY
fi

YOYO!

it basically means the code will run on almost all distribution

Days Ahead

Work
1) Change GUI
2) Add Functionality - Each user can train his face (no need to be root)
3) Add linear SVM Classifier
4) Use EHCI head tracker or Implement LK Inverse Compositional Tracker

-----------------------------------------------------------------------------------------------

InvERT , OH WE WON!

2009-01-29T04:09:00.000-08:00

Recent Update

AAM Deadlines have passed , but did not materialize , went to IIT Bombay techfest (for 4 days), for Invert and Connect, We won ( me, venkat (he is really good , seriously WAY to PRO) , yogesh ) INVERT, a reversing competition, level of problems was high, ( it gave me the motivation to be learn the art of reversing to advanced levels , this will be one of my subgoals this sememster), .also we won ON SPOT, (debugging and obfuscation Contest) , won a Circle 4 way scroll Mouse :P

Current Goals

-> Learning SVM in detail
-> Implement AAMs in OpenCV by Feb 2
-> Finish all the previous Goals
-> PFA (face authentication) first public full scale,official, unvieling (not the right word) during QUARK 09 BITS PILANI -goa campus techfest in the CS/IS department Stalls Feb 8th

My proposal for handwriting recognition project for SUN Code For Freedom 2009 got accepted
-> http://skewed.dev.java.net/

Right now thinking of going at it step by step.

Interpreting LK Algorithm

2009-01-12T12:52:00.000-08:00

Paper 1
Lucas-Kanade 20 Years On: A Unifying Framework: Part 1
Simon Baker and Ian Matthews

I will be updating the text here, after going through the paper

Motivation, Deadlines

2009-01-12T05:21:00.000-08:00

I want to release the next pam-face-authentication by Jan End,
For that i was looking at implementing some cool algorithms
Following are the deadlines i am setting for myself
0) Learn how AAM internally works ,and how to implement on top of OpenCV
1) Implement Person Specific AAM Module by Jan 15
2) Learn what Gaussian Mixture models are , and implement a GMM based Classifier by Jan 20
3) Do some feature "Engineering" , Figure out a fusion of featurs(dct mod2,gabor etc) Jan 22
4) Fix Distro Issue's
5) Fix Xauthority Problem
=> Basically include xauth.h use its api's
6) Implementing the same way how passwd utility works
=> this will make the application a lot more useful, each user can train his face without root priveleges
=> Storing "User" specific models in $USER_HOME
7) Build a proper documentation
8) Invite contributors to join the Project

9) Release it by JAN 26th

Learning Computer Vision: Registration Registration Registration

2009-01-12T04:52:00.001-08:00

Ok You would ask whats in the Title "Registration Registration Registration"
According to Simon Lucey ,phd from CMU , this is what his boss says Takeo Kanade ( from the LK algorithm)
When i started out on Computer Vision (face recognition) , CMU's RI website was like a gold mine of Ideas, and a world of amazing COOL outoff this world Application, Atleast stuff related with Face tracking,watch the videos posted on their site, its Jaw Dropping !!

Videolectures.net is just amazing, at home i had a 2mbps connection was able to watch most videos related to CompVi and ML,

One by Simon Lucey - on face registration and tracking is an eye opener, you could say OMG , SO KOOL,!!
David Mackay's - Gaussian Process is just wonderful! although i am struggling with understanding
it, might take a while

Carl Edward Rassmeusen , on Baysian inference and Gaussian Process! - well the Baysian Infernence was wonderful, havnt yet watched the Gaussian Part ( campus network sucks, need to go back home)

Andrew Blake:CV stuff, Amazing ( was obvious ) !!

Further Research Motivation for me
Simon lucey said in his lecture how AAM's are not good for varied appearance, and the PCA model does not Generalize well for all faces. He further explained about Local Model approach and how to speed it up, i was wondering if there are any othe features that we can use for appearance generalization.. hmmm!

Love for Machine Learning

2009-01-12T04:35:00.000-08:00

I havent been blogging much after mid of last semester, Last semester was just crazy!!

during the December holidays (around christmas) , i had went to CDAC-Kharghar Mumbai , to attend a Machine Learning Workshop.
The whole experience was amazing

It Kicked off , with Dr. Sasi Kumar (research Director) taking lecture on motivations for machine learning., anyways my motivation was for Image registration.
The Talk consisted of brief introduction on What features were , and about how to get most of out your machine learning algorithm - "Feature Engineering" , A lot of machine learning terms were Introduced.

Then Came Talks on Clustering and Decision Tree learning..

Followed by the Amazing lecure by Shekar on Neural Networks, he was really successful in giving the intuitive feel on how the the Training and Classification works on a NN.

After that we were taught SVM , by a lecturer whose name i forgot.
He did a fair job at it, but i did not learn anything new in it, It was meant as a introduction only and he at most part stayed away from optimization blocks of the algorithm

After all this final session was taken by Dr Sasi Kumar who explained why Feature "Engineering" and the Right Mix(fusion,ensemble) learning could do wonders and why we should stay away from Over fitting
he ended with "if it works 100%, its not Artificial Intelligence "

CDAC - Centre for Development for Advanced Computing - A government Institution is one of the best of its kind. The People were friendly, and if you stay at CDAC for a couple of days , you might never want to leave the place :)

-----------------

Face Detection , a new idea...

2008-10-22T11:49:00.000-07:00

I had read quite an Interesting paper on Face Detection during the summer,
it had applied Integral and Variance Projection on the entire image , and tried matching the Integral/Variance Projection Data ( found using Haar Face Detection) and matching it with RANSAC , why not try with Image on which LBP and angle Map operators are applied...

Tshirt for Code aka Code for Cotton

2008-10-22T11:43:00.000-07:00

Moments ago...
i received my Google Summer of Code T-shirt and Certificate...
Thanks Lesli, and the entire Google open source team. you guys rock..!

Bald ! Beaten and Bruised

2008-09-25T14:46:00.000-07:00

yeah its true

i have shaved off my hair

exams are OVER !!!!

aha :)

back to work on face recognition ...

Awesome Lyrics!

2008-09-11T11:00:00.000-07:00

Take my love, take my land
Take me where I cannot stand
I don't care, I'm still free
You can't take the sky from me.

Take me out to the black
Tell them I ain't comin' back
Burn the land and boil the sea
You can't take the sky from me.

Leave the men where they lay
They'll never see another day
Lost my soul, lost my dream
You can't take the sky from me.

I feel the black reaching out
I hear its song without a doubt
I still hear and I still see
That you can't take the sky from me.

Lost my love, lost my land
Lost the last place I could stand
There's no place I can be
Since I've found Serenity

And you can't take the sky from me.

aha! I am back ?

2008-09-10T11:51:00.000-07:00

Let us begin...

started out with Open source Development Team , Centre for software development
Organizing Software Freedom Day in Bits Pilani Goa Campus at 27 and 28 th of this Month ( yipeee )
Sun Campus Ambassador Program

Election for Students Council Bits Pilani Goa Campus

6 Courses this semester
Artficial Intelligence
Data Structures and Algorithms
Object Oriented Programming
Optimization
Computer Organization and Architecture
Operating Systems

3 Birthday Treats

Preparation for ACM ICPC

... and finally Test 1 for this Semester(aahh! run hide)

PAM -FACE-AUTHENTICATION 0.2

2008-08-17T12:40:00.000-07:00

You get the latest release from
http://pam-face-authentication.googlecode.com/

Video Demos
Demo 1

Demo 2

It uses Local Binary Pattern For Face Authentication , more information at

http://code.google.com/p/pam-face-authentication/w/list

THANKS A GAZZILLION to Nickolay Shmyrev ,Gnome

LBP -Face Recognition

2008-08-12T13:07:00.000-07:00

PAM_FACE_AUTHENTICATION

2008-08-12T12:45:00.000-07:00

Finally it is here

Download it from SVN at http://pam-face-authentication.googlecode.com/svn/trunk/pam_face_authentication/

Requires
OpenCV 1.0.0
Gnu Scientific Library 1.9 ( or greater)
libpam-devel (>= 0.99.8)

Algorithm For Authentication
Local Binary Patters (LBP)
Eigenface(not used any more)

Preprocessing
Roll Invariant
Homomorphic Filtering
Discrete Wavelet Transform Based Illumination Invariance( HE and Scaling in DWT Space)

Checkout Wiki @
http://code.google.com/p/pam-face-authentication/w/list
for Screenshots and More Information

Work to be done
Finish a proper config and makefile

Face Feature Tracking

2008-07-18T20:01:00.001-07:00

Here is the video

I will blog about this later .. really sleepy now

Problem with the pam module

2008-07-13T01:50:00.000-07:00

This is the Code from PAM , for authentication

PAM_EXTERN
int pam_sm_authenticate(pam_handle_t *pamh,int flags,int argc
,const char **argv)
{

ipcStart(); // Creates two blocks of shared memory one with 230400 bytes and another with 2 bytes

system("/home/darksid3hack0r/gtk-facetracker &"); // RUNS THE TRACKER , tracker contains authenticate and cancel button

if(startTracker(userTemp)=='y') // start tracker grabs frame from the cam and puts the image in the 230400 bytes created by IpcStart(); and Checks in the 2bytes block if any button is pressed

{ // printf(" y %s",user);
return PAM_SUCCESS;

}
else
{

return PAM_AUTH_ERR;
}
}

gtk-facetracker -reads from the 230400 block and changes the value of the 2byte block if authenticate or cancel is clicked

this works fine with ANYthing like su etc but Not with GDM

-----------------------------------------------------------------------------------------

with GDM , configuration of the pam-module , doesnt show the gtk-facetracker
in the log it says "client 4 rejected from local host AuthName: MIT-MAGIC-COOKIE-1 ID: =-1"

--------------------------------------------------------------------------------------------
i tested a pam Module with gdm which has the following code

PAM_EXTERN
int pam_sm_authenticate(pam_handle_t *pamh,int flags,int argc
,const char **argv)
{

return PAM_AUTH_ERR;
}

-------------------------------------------------------------------------
and this works , a window is shown
------------------------------------------------------------------------

if i configure this with anything like su or sudo
i get the an error which points to -> www.gtk.org/setuid.html

---------------------------------------------------------------------------

if i put this code in any binary and run that binary with PAM module , the same "client 4... " error shows up in log and the window is not displayed

Code:-

Display *d;
int s;
Window w;
XEvent e;

/* open connection with the server */
d=XOpenDisplay(NULL);
if(d==NULL) {
printf("Cannot open display\n");
exit(1);
}

s=DefaultScreen(d);

/* create window */
w=XCreateSimpleWindow(d, RootWindow(d, s), 10, 10, 100, 100, 1,
BlackPixel(d, s), WhitePixel(d, s));

/* select kind of events we are interested in */
XSelectInput(d, w, ExposureMask | KeyPressMask);

/* map (show) the window */
XMapWindow(d, w);

/* event loop */
while(1) {
XNextEvent(d, &e);
/* draw or redraw the window */
if(e.type==Expose) {
XFillRectangle(d, w, DefaultGC(d, s), 20, 20, 10, 10);
XDrawString(d, w, DefaultGC(d, s), 50, 50, "Hello, World!",strlen("Hello, World!"));
}
/* exit on key press */
if(e.type==KeyPress)
break;
}

/* close connection to server */
XCloseDisplay(d);

----------------------------------------------------------------------------

Life with GDM and X Server and some other components

2008-07-12T06:54:00.000-07:00

This week , i configured the PAM module with gdm, it did not work intially. After searching
for a solution, finally realized that the GUI was not able to connect to the X Server because it did not have the x auth MIT-MAGIC-COOKIE-1 ,

When GDM daemon runs ,it also runs the GDMgreeter with gdm:gdm , it also uses Xauth cookie for connecting to the xserver
and this is stored atleast for openSUSE 10.3 at location /var/lib/gdm/

and when PAM Module executes the Helper GUI binary ,it is run as root and it checks the XAUTHORITY environment variable for the location of the cookie file, which contains a different cookie

so i added the cookie hex to user roots xauth list for display :0

i tried again, it did not show the gui
{ i am not sure why the reason is not clear and have to found out

[ maybe related with threads its not able to start a gui binary ]

if this can be fixed , problem is solved to greater extent

}

then i changed the PAM module code to include the displaying of a GTK window which i had changed to a separate Binary because when i was testing with SU , it threw an error and pointed to this http://www.gtk.org/setuid.html

and when i tried this it worked!!!!

now next stop was with the xlock , this however when i tried with executing the binary. no binary showed up , but then when i entered the normal password and went in to the desktop i saw the binary gui window , which means the Xlock was blocking the Gui window

i tried gui code with pam with Xlock and it did not work , maybe because of the earlier pointed error http://www.gtk.org/setuid.html

Some questions that needs to be answered Divide PAM into two PAM modules one for GDM and another for setuid functions (su sudo etc) and forget about Xlock

or get the execution of the binary work with Xlock (binary is being executed but its blocked by the xlock screen) + anothe problem is the password textbox still shows

and get the execution of the binary work with gdm

A Face based Authentication Story

2008-07-06T14:50:00.000-07:00

After going through a lot of design change , the project now has come alive. Current way of doing things seem to work fine :) and Also after experimenting with a lot of ipc's , finally i have decided to go with Shared Memory. i also made two core libraries one for detection another for configuration , the problem of code getting repeated have been solved.
There is bunch of functions which was written last month but not have been used , for eg the eye detector , the homographic filtering code , wavelet based illumination control code etc. Now these three things have to be done :-

Finish Up UI & Bug Testing ( have add a little more code have to added )
Add functions onto the detector and configuration libraries
package and deploy

The following are the screen shots the project , it will take another 15 days for finishing the first task completely and the second task [ being a bit optimistic here ]

Hakunamatata ... its a wonderful phrase

2008-06-25T16:09:00.000-07:00

Yes , dear readers , HAKUNAMATATA .... yes it means no worries... for the rest of the day ... and probably why i did not go mad the past two days... i was working with dbus , finally got the low level api code to send arrays of Picture data at 15fps... [ well good accomplishment ] BUT then again CPU usage 80-90% on each core... WT* , and the part where i was receiving data had a memory Leak....... WT* , then again.. i had to find something else.. ya and SHMGET came to the rescue.. shared memory rocks!!! walaaah! and the cpu usage is around 35 % each and memory usage is also constant , no leaks... :P... now i can say yipeee and concentrate on the UI

Problem on Connected Components : Can you Solve it?

2008-06-18T15:27:00.000-07:00

Problem
Q1 , you are given a N X M board which can contain any of the two values {1 , 0} , and a cluster is defined as a conncected component containing 1's either horizontally vertically or diagonally , basically any of the 8 positions possible near by to a point (x,y) ... now How Many Connected Components are MAXIMUM possible?

i need the answer in N and M ?

my bad , i previously wrote "cluster" instead for Connected Component

Plan for this week

2008-06-18T02:49:00.000-07:00

what should i do this week?
my options are
1) Work on vpf and eye tracking
2) Work on Preprocessing (homomorphic filtering)
3) Work on User Interface?

GSOC UPDATE , YiPEEE

2008-06-17T00:06:00.000-07:00

Yipee!!!
This my latest blog video , Check it out , results are too good .. i couldnt test it on a lot of people. i sent it to Nikloy Shmyrev , his results were good too , now i have to force myself to sleep...!!!

GSOC Update [ breakthrough ]

2008-06-14T18:40:00.000-07:00

The Eyetracker algo which i developed last time IS NOT UPTO THE MARK , or also KNOWN as it SUCKS
i read through a lot of papers ( thank you google scholar) , finally sat down and started writing some code
Some Interesting Observation of its VPF , couldnt get time to plot the horizontal , but probably it should look the same. Does anyone know how to use Any good Graph Plotter ? currently i am using openoffice .anyway this is a major break through , though i will have put some time in it to get a good tracker out of it and Hope that it works with everybody's eye. It s time for me to sleep , and i have to prepare a report of what i have done till now tommrow... o h wait i mean today
these are the results

...............
zzzz.......

2008-06-12T08:51:00.000-07:00

Last 11 days was the most productive time ever , it started with Increasing accuracy for recognition , so i had to work on LDA that Linear Discriminant Analysis , dimension reducing technique which creates new axis in the space such that the it maximizes the (interclass variance/intra class variance) , if you solve this you will get something like S(w)x=(lamda)S(b)x now this eigen value problem can only be solved if S(b) has an inverse which reduces the equation to a normal problem, but in my case it wont have one, so i had to research a bit and finally ended up with Jeiping Ye's paper which you can obtain here
http://www.cs.umn.edu/~jieping/UOLDA
i went through his paper , took me 4 days to understand it , then i looked at his matlab implementation , which i found had some wrong statements , so i got really confused , since i wasnt sure that those were wrong , i mailed and he replied pretty soon. then i had to write C program for doing this , computing SVD , yeah there were a lot of resources , i got stuck at computing rank and then karan helped me remember how to calculate as we did in college , then i wrote that implementation also. then seeing that it was not going anywhere i dropped it and kept it for future ,
now secondly accuracy at different illumination , there were methods like HE(histogram equalization) which i am currently using, then there was HS , and something which starts with letter G , hehe , i forgot what that was , now i finally ended up at an AWESOME paper which CLAIMS 15% increase in recognition under varied ILLUMINATION
http://www.image2003.com/paperen/open.asp?ID=10550
so i went through the paper , had to figure out how DWT works and finally i got it ( this made me wonder ... should i have taken an Electrical Engineering?) now i had to implement DWT , which is a MIGHTY task for beginner like me , so i searched the net for good libraries , finally found 2 , but BOTH >5 mb , and i just wanted to use just one function ... [ so again i am keeping this for later ]
finally the EYE , [ i used normal 10th standard physics and some image processing techniques and here are the results :D ]

gsoc Project update

2008-06-01T05:45:00.000-07:00

finally i am done with all the modules
pam_face_authentication
facemanager
facetracker [.so ]
faceauthenticate [ .so ]

Integration Problems Faced:
Recognition Algorithm was not working (! huge disaster)
FIXED: by adding more data[faces] , i was testing with just 2 pics

MAJOR PROBLEM:
i tested my pam module on su
it works fine if i do su root when i have already set myself as root
but if i do su root from something other than root
i get the following error
Please Look at the Center of the Screen for the next five seconds.

(process:17665): Gtk-WARNING **: This process is currently running setuid or setgid.
This is not a supported use of GTK+. You must create a helper
program instead. For further details, see:

http://www.gtk.org/setuid.html

FIXED: i disable the The Face Tracker GUI and it works fine.

DESIGN CHANGE
To incorporate the GUI

system out a GUI app [ i dont think setuid problem will occur ]
then use DBUS to send picture data and tracker data to the GUI tracker
will work? no idea . I will have to discuss this with my mentor
till then ciao

This was my Deadlines
2.1 PAM Implementation May 26 - July 6 [almost done]
2.2 Implement Face Detector & Recognition Algorithm June 1 - June 12 [ done ]
2.3 Database Design & Working Out Access Rights June 12 - June 18th [ not using a database any more , just a normal file :) ]
2.4 Face Manager June 19 July 5 [ basic done]
2.5 UI For GDM June 25 July 10 [ NOT AT ALL DONE ]

UI : i will keep it for later
next 10 days , i will work on a plugin based architecture for recognition and detection algorithms
add more arguments in facemanager to have total control over the process
and implement some more recognition algorithm

New Theme and Gsoc Update

2008-05-31T03:23:00.000-07:00

Check Out this THEME yo'aal!!! i like it very much ..
now back to important stuff

Google Summer of Code [ Milestone 1 ] - Progress
facetracker completed !
facemanager completed !

To Do
faceauthenticate Library
pam_face_authenticate
and Finally Integration

Now, back to work , i will update this blog in the night

Google Summer of Code Update and other Updates

2008-05-28T12:56:00.000-07:00

After researching on various possible implementation, my mentor and i finally came up with this design,
main problem faced was the interprocess communication that was needed , this was new to me, my mentor suggested d-bus , after researching on it for sometime , another implementation struck me, how about forking and using pipes but here also i was not successful , because pam modules are implementation of library functions , so finally system() came to the rescue , but my mentor gave me another idea , how about system() the face_tracker and the dlopen() the face_authenticate library which is much cooler .

First Milestone [Command Line Implementation of the Following Modules]

pam_face_authentication
facemanager or facemgr [ comment ]
face_tracker
face_authenticate

----------------------------------------------------------------------------------------------------------------------------------
In other news

localstopten.com to be finished this sunday
quit uinterns.com [ something had to go , no time for everything ]
Practice School at Lasersoft Infosystems going well

-----------------------------------------------------------------

Almost over

2008-05-05T07:34:00.000-07:00

i got in google summer of code , my mentor is Alex Lau, he is a pro linux developer at Novell, China
http://code.google.com/soc/2008/suse/about.html
and guess what
My elder brother Also got in at Apache Foundation
http://code.google.com/soc/asf/about.html ( check for Robin Anil )
now that is JUST AWESOME,

Final Sem exams are up
2 over :)
5 to go

i will be back to the blog on 13th

Time Draws Near

2008-04-21T08:11:00.002-07:00

GSOC results are going to be out in the next 4 hours and i am feeling really nervous and tensed.

2008-04-21T08:11:00.001-07:00

GSOC results are going to be out in the next 4 hours and i am feeling really nervous and tensed.

Gsoc Selections

2008-04-19T06:20:00.000-07:00

Gsoc Selection Results are to be out on 21st, got a mentor in OpenSuse for face authentication, but it now all depends on luck whether am i above the above line or not , hoping i make it

uinterns.com, vipashyin.com , talkofcampus.com and localstopten.com

2008-04-18T02:55:00.000-07:00

4 websites !
this month
uinterns.com a venture by 6 bitsians (inc me) , the right place to look for internships
localstopten.com , alex and chester contacted us on 15 for making their website, now the final phase is in progress
vipashyin.com - harsha's venture, the best web development team providing custom web apps!
talkofcampus.com - brainchild of harsha, a online place for students to get together..

PS result , the good ,the bad and the ugly....

2008-04-05T09:59:00.000-07:00

i got my Practice School Allotted at LaserSoft Chennai, had my Quantum Computing EXAM in which i got a COMPLETE black out, m getting a ZERO in the paper, life was down and ugly till me teesta and anuraag went and played caroms,talked about the random stuff , had good dinner. Now i sit here typing this blog, i have my next exam on monday.Need to ace it! GSoc Submissions are almost over,waiting to get selected so that i can have something to look forward to this summer!

Google Summer Of Code Proposals

2008-04-01T15:34:00.000-07:00

Well, this is my second post after making this blog, i am not much of a blogger.
First Proposal Was to make a Face based Authentication
After some googling/research i narrowed down the components i have to use, first was something called PAM - Pluggable Authentication Module , just PAM'fy the Code i will be writing and i will get out of the box authentication for any purpose , just need to add some few lines to the config script
and yeah EigenFace again.. i will be using this for face recognition and the time tested viola and jones method for detection.
Second was gesture based Navigation
Viola & Jones Method for face can work out for any objects , i just have to create samples images and train the classifier and it would let me detect any kind of object that i wish , so why not train to recognize different gestures and see which gesture is being detected and do the suitable action. I recall telling the professor who i had contacted for further research into this, he told me to make a module for image correlation and if i am successful he said we can continue from that .. how about it .. a paper on Gesture Based Navigation in HCI .. totally awesome!!!!!
and if things go according to plan _______________ << reason for this is that i donot want to jinx it by saying it :D
o wait... i have my tests this week , need to score above Average !!1
well blogging seems cool.. this is my first sane blog ..

My First Post

2008-01-26T03:08:00.000-08:00

Human Mind : a mysterious system
The greatest achievement for a human would be to understand the BRAIN . it works 24/7 till the end taking(no breaks). Why do i do something, what is controlling me do it, am i responding to some stimuli? why did i just create a blog? why did i not create it 2 years ago ? why i am i thinking what i am thinking right now? there are infinite questions and no answers for any them!

Why did i just write a paragraph about the brain? Is it that i want somebody to read it and see how i think? Does humans have the urge to show other fellow humans what they think like? Why am i writing humans so often? Is it because i have watched a lot of SciFi Alien Humans Involved KickASS! movies? Is it because i am an alien? How do i prove to myself that i am not? Maybe I am an alien Who have transformed to a human and must have erased all alien traits that even i do not know whether i am or not or is it all Because i have watched TOTAL RECALL??? ...

Given infinite Computational resources can we humans Figure out the human brain? Whenever i see some mathematical equations producing results which are quite counter intuitive but REAL , i think to myself does brain Work like one of these equation so counter intuitive that we donot have the computational resource in our brain or as hardware to figure out what it is?

The idea of Diagonalization in theory of computation, even though i still did not get the intuition for it, there is some part of my mind telling me that the idea is flawed or Can it be used to explain why there is functions that we cannot make , or is the idea Flawed Because of the definition of infinity is flawed?