Fig 1.
The distortion frequency (x-axis) represents the frequency of cyclic variation. The severity of distortion (y-axis) is the magnitude of displacement. This creates changes in the configural and featural information in a face.
Fig 2.
Matching image chrominance and luminance.
Images of different people were selected at random from the IASL Database. A histogram of luminance is shown below each image. Pixels from image 1 and 2 were sorted in ascending order. The mean image was created with the combined/averaged luminance and chrominance information, and this was reapplied onto the original image to create an equalized image.
Fig 3.
Upright and Inverted Face ID Task displayed the undistorted target face on the left and 4 distorted faces centered on the screen including the matching face and 3 distractor faces. Participants were prompted to choose the face that matched the identity of the person displayed on the left.
Fig 4.
The X-axis displays the Distortion Frequency in Cycles per Image (CPI), and the Y-axis displays the Tolerance to Distortion in wavelengths. Participants are color coded, with the average tuning curve shown in black. 4a) Shows the unique average tuning curve for the Upright Identification Task. 4b) Is the Inverted Identification Task. 4c) Displays the difference between the upright and inverted tasks, representing the Face Inversion Effect (FIE). The FIE is absent at the middle frequencies.