Smile for the camera!


Christian Möller (DE) (US)
Sean Crowe (US)

Happy faces are everywhere—in commercials, on billboards, in Hollywood movies. “Cheese” provides a glimpse behind the mask of this seemingly ubiquitous joviality.

The impetus to produce “Cheese” came from the relentlessly friendly smiling faces of the entertainment industry. It all started with an ad in a Hollywood trade paper: “Actress wanted, TV anchorwoman type, for video portraits.” More than 800 young actresses responded.

Before a rolling camera, six actresses tried to smile as long as possible—up to an hour and a half. Each of these extended smiles was simultaneously monitored by a computerized perception system; as soon as the acted-out friendliness sank below a certain level, an alarm signal was triggered as a means of prompting the actress to put a bit more sincerity into her performance.

“Cheese” is a staging of a form of human-computer interaction in which the computer plays the lead part. The actresses’ very real emotional uneasiness—which they are able to conceal by means of convincing jobs of acting—comes across only during the breaks that are necessary to relax their strained facial features.

This work is based on research on emotion recognition being done at the Machine Perception Laboratories of the University of California, San Diego.

Models: Melissa Berger (USA), Laura Clumeck (USA), Natasha Desai (USA), Kyra Locke (USA), Susan Marshall (USA), Cameo Cara Martine (USA)

Special thanks to Pierre Moreels (F), Pietro Perona (I), Javier Movellan (ESP), Marni Bartlett (USA), CALTECH – the California Institute of Technology, the Center for Neuromorphic Systems Engineering in Pasadena and the Machine Perception Laboratory of the Institute for Neural Computation at the University of California, San Diego.