Face Substitution

source: vimeo

Working on an idea with Arturo Castro vimeo.com/29279198 I feel like “good” blending looks almost too natural to be surprising. It doesn’t leave any interpolation up to your imagination. It’s possible to push this style further, so it’s less of a blend and more of a replacement, but then you get unnatural colors and shadows.
FaceTracker library from Jason Saragih web.mac.com/jsaragih/FaceTracker/FaceTracker.html
ofxFaceTracker addon github.com/kylemcdonald/ofxFaceTracker
openFrameworks openframeworks.cc/
Download the app here github.com/arturoc/FaceSubstitution
.
.
.
.
.
.
source: kylemcdonaldnet

Kyle McDonald is a media artist who works with code, with a background in philosophy and computer science. He creates intricate systems with playful realizations, sharing the source and challenging others to create and contribute. Kyle is a regular collaborator on arts-engineering initiatives such as openFrameworks, having developed a number of extensions which provide connectivity to powerful image processing and computer vision libraries. For the past few years, Kyle has applied these techniques to problems in 3D sensing, for interaction and visualization, starting with structured light techniques, and later the Kinect. Kyle’s work ranges from hyper-formal glitch experiments to tactical and interrogative installations and performance. He was recently Guest Researcher in residence at the Yamaguchi Center for Arts and Media, Japan, and is currently adjunct professor at ITP.
.
.
.
.
.
.
source: creativeapplicationsnet

It all kicked off about 2 months ago when Kyle McDonald posted ‘FaceOSC’, a tool for prototyping face-based interaction. The oF wrapper Kyle created is built on non-commercial open source FaceTracker code from Jason Saragih. Kyle included an example oF app with the ofxFaceTracker addon for openFrameworks which can be downloaded here.

Then few days ago, Arturo Castro carried on the work by exploring different face substitution techniques. Using the same library, mesh obtained from a photo is matched to his own face in the video. Applying some color interpolation algorithms from Kevin Atkinson’s image clone code: methodart.blogspot.com/​ gives it the blending effect that can be seen in the final footage.

Now only few minutes ago, Kyle posted this video, named “Scramble Suit” inspired by fictional technology from Philip K. Dick’s 1977 novel, “A Scanner Darkly”. It’s effectively a cloak that hides the identify of the wearer by making it impossible to describe or remember them. Kyle points to an excerpt here.

Very exciting bunch of face tracking experiments. Looking forward to see what come next…
.
.
.
.
.
.
source: storyitdacokr

Face Tracking은 Kyle McDonald에 의해 개발된 얼굴 기반의 상호작용 도구라고 한다.
사진에서 얻은 메쉬 정보를 활용해 다른 얼굴로 대체하는 기법이라고 하는데, 이 기술로 무엇을 할 수 있는지, 이 기술이 궁극적으로 어디에 도달할 것인지 정해진 바는 없다. 아직은 열정적인 프래그래머가 흥미진진하게 몰입할 수 있는 장난감 같은 것이지만, 흥미로운 건 사실이다.

이 영상을 보고 있자니 프란시스 베이컨의 인물화를 보고 있는 듯한 기분도 든다.
카일은 Face Tracking을 활용해 필립 K. 딕의 소설 “스캐너 다클리”에 나오는 ‘Scramble Suit’이라는 가상 기술로부터 영감을 받은 영상을 공개했다.

Face Tracking, 이 기술이 우리에게 또 어떤 새로운 경험을 선사할까?
관심을 갖고 지켜보자.