The link for stating applications and patches to emulate Moeller’s Cheese Installation is:
http://www.processing-plant.com/FaceOSCMavericks.zip (if you run Mavericks)
If you do not run mavericks, you need another version of the face feature detection.
You can find the app under: https://github.com/downloads/kylemcdonald/ofxFaceTracker/FaceOSC.zip
Simply run this face app instead of the Mavericks version.
Cheese is described under: http://oss.adm.ntu.edu.sg/2014-dm2000/christian-moeller-cheese/
The FaceOSC patch will show you the different features extracted and their values. In order to determine if a person is smiling, both smile height and width are needed. First observe these values and see how they match a style of smiling. Second, program some logic that will enable you to detect various kinds of smile. Is the length the only factor or both length and height compounded? Third, have a feedback displayed for a smile or not (red/green). Fourth, start playing with sound…and trigger an annoying alarm that shuts down only when people smile. Five, imagine a different gesture tracking, i.e., have the person whistle instead, raise eyebrows in sequence, etc.
This emulation introduces a few new items: sound playback and inter-application communication. Due date for delivery is TBA since a few more concepts need to be covered in class.
all the best, lpd