About
Face recognition that blends Marilyn Monroe’s face into yours.

How It Works

Step 1 : Open Camera & Face Position
As usual, we need to grab the camera and track the face’s position to get things started. This part of the patch receives a bounding box (e.g. the message ’79. 122. 178. 221.’) that places the image. The t b l object stands for ‘Trigger, Bang, List’.


Step 2 : Import & Place Image
Next, an image needs to be imported with a black background (example shown below). Image has to be full face frontal. Calculation needs to be made to align the image onto face.

jit.fastblur object is to blur the image. Instead of using the fourth mode (square), I used the first mode (column) because I wanted my image to not get blurred completely.

Step 3 : Blend Them Together!
Lastly, we use the blurred image earlier as an alpha channel. @planemap is an alpha channel based on image. The four values represent ‘Opacity, R, G, B’. I set the opacity quite high as I wanted the details of my image to show more.

!pass = take whatever is there (use image as alpha channel)
pass = go straight

Challenges
I had trouble trying to get the image to fit on the face nicely. I was suspecting that the bounding box is making my image to get squashed vertically (the top part stops in the middle of my forehead while the bottom part stops right under my mouth).

The Interaction