Pointing'04 ICPR Workshop
Cambridge, United Kingdom - 22 August 2004

Head Pose Image Database



    1. Image Database :


            The head pose database consists of 15 sets of images. Each set contains of 2 series of 93 images of the same person at different poses. There are 15 people in the database, wearing glasses or not and having various skin color. The pose, or head orientation is determined by 2 angles (h,v), which varies from -90 degrees to +90 degrees. Here is a sample of a serie :

sample


            Each file tar.gz contains one serie of 93 images and has an approximative size of 1MB. For example, the file Person05-2.tar.gz is the second serie of the person number 5. All images are in JPEG format. There are a few examples of series in MPEG format. The file Front.tar.gz consists of 30 frontal images of persons of the database. This serie is useful to learn or to test on frontal images.


            Filenames are constructed according the following grammar :

personne[PersonID][Serie][Number][VerticalAngle][ HorizontalAngle].jpg

where :

    PersonID = {01, ..., 15}:
                stands for the number of the person.

    Serie =  {1, 2}
                stands for the number of the serie.

    Number = {00, 01, ..., 92}
                the number of the file in the directory.

    VerticalAngle = {-90, -60, -30, -15, 0, +15, +30, +60, +90}

    HorizontalAngle = {-90, -75, -60, -45, -30, -15, 0, +15, +30, +45, +60, +75, +90}


For example, file personne08123-30+45.jpg :

    PersonID = 08
    Serie = 1
    Number = 23
    VerticalAngle = -30
    HorizontalAngle = +45
 


Negative values
Positive values
Vertical Angle
Bottom
Top
Horizontal Angle
Left
Right


In case the vertical angle is -90 or +90, the person is looking at the bottom or the top, and then the horizontal angle is 0. Each serie contains therefore 7 x 13 + 2 x 1 = 93 images.



    2. Image Acquisition :


            All images have been taken using the FAME Platform of the PRIMA Team in INRIA Rhone-Alpes. To obtain different poses, we have put markers in the whole room. Each marker corresponds to a pose (h,v). Post-it are used as markers. The whole set of post-it covers a half-sphere in front of the person.

            In order to obtain the face in the center of the image, the person is asked to adjust the chair to see the device in front of him. After this initialization phase, we ask the person to stare successively at 93 post-it notes, without moving his eyes. This second phase just takes a few minutes. All images are obtained by using this method.
           

View
Top sight



monica cote
Side sight



Fame

The FAME Platform



    3. Downloads :


            Here are  images sequences and example videos downloads :


Videos :


Video-Person01                         Video-Person03                         Video-Person08



Image sequences :


Person01-1                                         Person01-2
Person02-1                                         Person02-2
Person03-1                                         Person03-2
Person04-1                                         Person04-2
Person05-1                                         Person05-2
Person06-1                                         Person06-2
Person07-1                                         Person07-2
Person08-1                                         Person08-2
Person09-1                                         Person09-2
Person10-1                                         Person10-2
Person11-1                                         Person11-2
Person12-1                                         Person12-2
Person13-1                                         Person13-2
Person14-1                                         Person14-2
Person15-1                                         Person15-2
Front




This database can be used for any purpose, provided that the following article is cited :


N. Gourier, D. Hall, J. L. Crowley
Estimating Face Orientation from Robust Detection of Salient Facial Features
Proceedings of Pointing 2004, ICPR, International Workshop on Visual Observation of Deictic Gestures, Cambridge, UK



The author can be contacted for details not mentioned here:
Nicolas Gourier (nicolas dot gourier at inrialpes dot fr)