Mona Lisa's smile

Find out the true feelings behind Mona Lisa's smile.

I am sure this application has value but honestly I do not need a computer to tell me my wife is upset, sad, or happy, most of the time I can figure that out for myself. But none the less I found this article posted on Technology news today interesting. In short, a number of academics from various universities have a developed face tracking software that can tell if the person is happy, sad, afraid, or angry. They do this by creating a virtual wire frame around the persons face and measure the differences against known neutral. To illustrate the idea and I am sure also to build PR they have used the famous painting of Mona Lisa as an example. According to the test Mona Lisa is happy, slightly dugusted, a little afraid, and a little angry. I think we need to remember when reading the statistics below that the painting was created in the early 1500’s, this was not taken with a digital camera. Mona Lisa sat their probably a few times for many hours not knowing if Leonardo DaVinci was doing a good job or not, but hey she was a model for this great painter who wouldn’t be happy.

Read more details below.

The Mona Lisa's Smile

Face-tracking software details emotions behind art’s most famous painting, the Mona Lisa.

Professor Nicu Sebe of the University of Amsterdam says he can calibrate those emotions using face-tracking software developed in collaboration with Professor Tom Huang of the Beckman Institute for Advanced Science and Technology. Sebe, Huang, and Huang’s students at the University of Illinois at Urbana-Champaign have spent years writing algorithms that quantify facial expressions and the emotions they represent into a face-tracking software program. Recently, for fun, Sebe decided to apply the program to an image of the Mona Lisa, painted by Leonardo DaVinci probably between 1503 and 1506. The subject of the Mona Lisa is thought to be Lisa Gherardini, wife of a cloth merchant from Florence, Italy. But her true identity and other aspects of the painting’s history are as mysterious as her smile.

Now, at least that part of the painting’s mystery can be explained. Sebe said he used the computer recognition software program to compute Mona Lisa’s emotional state. Sebe placed a virtual wireframe model first developed by Huang and his students over the image’s face and computed displacements from a prototypic, neutral Caucasian female face from the Cohn-Kanade database at Carnegie Mellon to obtain the results. The numbers showed a great degree of happiness (82.67 percent), along with lesser feelings of disgust (9.17), fear (5.81) and a tiny bit of anger (2.19). Sebe interprets the disgust percentage as possibly being a measure of irony (many viewers have sensed an ironical smile), but he is not sure how to interpret the fear angle.

While decoding the Mona Lisa may be fun, the human facial computer recognition project Huang and Sebe collaborated on has important potential applications, including business and educational uses. For example, Huang has been working on improving the computer interface experience for middle school students by using a face-tracking program to give the computer a better understanding of the user’s emotional state. The work on emotion analysis by Huang and his students was funded by the National Science Foundation and the Yamaha Motor Company.

The Beckman Institute for Advanced Science and Technology at the University of Illinois at Urbana-Champaign is an interdisciplinary research institute devoted to basic research in the physical sciences, computation, engineering, and biological, behavioral, and cognitive sciences.

The Beckman Institute for Advanced Science and Technology

Technorati Tags : face emotion software |