The phone will use the screen to mimic a camera flash for front-facing photos.
We cut up full-face photos.
They have spread after spread of erotic watch-face photos.
The array always contained 19 real face photos and 1 mask photo.
Human face is detected on the face photos using Viola Jones method.
This resulted in six variant image sets, each consisting of one mask photo embedded in 19 real face photos.
Experiment 1 (UK; n = 60) revealed no evidence for overt detection of hyper-realistic masks among real face photos, and little evidence of covert detection.
Starting with the 20 real face photos, we created different stimulus sets by substituting one mask for one real face of the same type (young male, old male, or old female).
On the detected faces and computer generated sketches, feature extraction is performed using supervised auto-encoder to build deep architecture and matching is performed between computer generated sketches and face photos using Parallel Convolutional Neural Network (PCNN).
A low proportion of viewers believed that they were in the Mask condition (Question 3) and a low proportion picked the mask out from an array of real face photos (Question 4).
The key contribution of this paper is we have developed a novel technique to match computer generated sketches with face photos and implemented it on ODROID XU4 single board computer which makes it feasible to be used in real-time.
