Knowledge Base

Export (Detailed Log And API)

Consumption Behavior Statistics

This option is only available if your FaceReader license includes the Consumption Behavior Module. Select this option to add the variables Intake count and Chew motion count to the Detailed log.

Gaze Fixation Index And Location

This option is only available if your FaceReader license includes the Advanced Research Module. Select this option to add three columns to the Detailed log file: Gaze Location X, Gaze Location Y and Fixation Index. Gaze Location X and Gaze Location Y are the x,y coordinates of the test participant's gaze point on the screen. The X,Y-coordinates are in pixels relative to the upper left corner of the screen. The Fixation Index is the total number of fixations up until that point in time.

2D Landmarks

With this option selected the X-Y-coordinates of the dots around the eyes, nose and mouth of the mesh are added to the log file. The log file has 134 numbers in the column Landmarks. The first two numbers are the X- and Y-coordinates of landmark 0, the third and fourth are the coordinates of landmark 1 etc.

The following landmarks are present in the detailed log:

• 0-9 – left eyebrow
• 10-19 – right eyebrow
• 20-21 – corners left eye
• 22-23 – corners right eye
• 24-26 – nose
• 27-66 – mouth

The X,Y-coordinates are in pixels relative to the upper left corner of the image.

Please contact Technical Support if you want to know the position of the landmark points in the face. They can send you a photo of a face with the 67 landmark point positions. They can also provide you with a photo of all the 468 landmark point positions in the face.

See Show Mesh in FaceReader's output

Log Complete Set Of 3D Landmarks

Select this option if you want to include the X,Y, Z coordinates of all the (468) landmark points in the Detailed log.

Please note that the 3D landmarks are not in pixels but in millimeters relative to the camera.

To project the 3D landmark points on the camera or video images you need the camera matrix:

C =
fx, 0, cx
0, fy, cy
0, 0, 1

if

w = image width in pixels
h = image height in pixels

then:

fx = 1.2 * Max(w,h)
fy = 1.2 * Max(w,h)
cx = w/2
cy = h/2

Then each 3D world coordinate (xw, yw, zw) can be projected to an image coordinate (xm, ym) by:

xm = (xw * fx)/zw + cx
ym = (yw * fy)/zw + cy

Source: FaceReader 10 Reference Manual (Help), Noldus Information Technology