Digital Skin/Face Paint, 2019-20
In my latest works in progress, I am exploring the interchangeability of digital and physical. I have always been fascinated with the in-between space of digital design and physical manifestation of the design. Implementing computer vision software, OpenCV, a digital image/portrait is aligned with and projected onto a face detected through face recognition software. A digital image as a skin on top of one’s physical face is also a playful and absurd act and response to how facial recognition has become an omnipresence technology in service of societal control.
Digital Face Paint (DFP) is a spatial augmented reality system that uses an RGB+depth camera and projector to project augmenting information onto a target surface. Specifically, DFP uses an Intel Real Sense SR300 camera to see and track a human subject’s face and, from the detected facial feature points, warps a template image to create the digital mask. This mask is further warped from the perspective of the camera to that of the projector and then projected onto and in registration with the subject's face. Because the SR300 camera works in the invisible NIR range, it does not see the projected image. So the projected image does not affect DFP’s ability to track the subject. Also because DFP sees the human subject’s eyes, it clips the eyes of the mask to avoid any irritation caused by looking directly into the projector.
This is a collaboration between School of Art and Visual Studies, Department of electrical and Computer Engineering and Department of Radiology at the University of Kentucky.
Siavash Tohidi, Dr. Daniel Lau, Dr. Michael Winkler.