Emotion Recognition Research


Starting this semester my team has begun looking at emotion recognition technology. This is a funded research project with one of our outside partners. The core technology is from a colleague’s team in computer science, building off open source research. Our team is looking at the creative potential and privacy implications. Every technology can be exploited creatively, and experimental takes on emerging technology often offer perspectives on social impacts that were not obvious during the technical research stages. 

What we hope to do is create an application that calls attention to this emerging technology and educates us a bit about how it might affect our lives in the near future.

As a side note, one of the things we’ve found initially is that all the training for the systems is done by volunteers basically making faces. The thing is, they can’t act. So what is happening, in effect, is that emotion recognition systems are being trained to detect fake emotions. My thought here would be to a) make people feel the emotions for real: make them laugh, etc. scare the sh*t out of them, or b) hire Daniel Day Lewis and have him go through all the emotions.