My son got one of the Emotiv headsets as a college graduation present. It’s pretty amazing tech, early-adopter gear, for sure, but with tons of potential. He’s already coming up with ideas for game designs and art installations controlled by it (he has a degree in game design). In this video, Robert Oschler, of Robots Rule, uses his Emorate software to demonstrate the power of “affective computing,” using computers to detect and react to human emotions. Here he uses various emotional responses to index, bookmark, and navigate a video using the Emotiv headset and Emorate.
2 thoughts on “Recording emotions with the Emotiv headset”
Comments are closed.
ADVERTISEMENT
Join Make: Community Today
Very nice – love to try out the Emotiv myself.
There are several different models of emotion, you have the categorical ones as here – which derive from Darwin’s early work – http://en.wikipedia.org/wiki/The_Expression_of_the_Emotions_in_Man_and_Animals – and latterly Ekman – http://en.wikipedia.org/wiki/Paul_Ekman – There’s much debate about what the basic set are, Ekman has six: Anger, Disgust, Fear, Happiness, Sadness and Surprise.
A dimensional model of arousal and valence is also commonly used – http://en.wikipedia.org/wiki/Emotion_and_memory#Emotional_arousal_and_valence – where each emotion is plotted on these axes. That’s of interest because simple and cheap sensors can measure Galvanic Skin Response (GSR), also known as Electrodermal Response (EDR) – http://en.wikipedia.org/wiki/Electrodermal_response – a good indication of arousal. Valence is trickier. So you can get a measure of how excited somebody is, but not whether this is a positive or negative emotion for them.
I did some experiments with GSR and video analysis back in 2006: http://homepage.mac.com/dave_chatting/portfolio/#affectivesocceranalysis – we were looking at fans watching soccer matches and finding where the exciting moments were – goals, penalties, etc.