Just a small reaction to the interesting discussion on language / interpretation. We are working in a research project aiming at facing "expressiveness" in movement (and in music). The project is funded by EU (IST) and is named MEGA (Multisensory Expressive Gesture Applications, www.megaproject.org). The focus is on "expressive gesture". In short, one of our works consists of analysing the same movement (that we call "microdance", a movement sufficiently long, a phrase, in which it is meaningful to inject expressiveness by a dancer), performed by the same or different dancers with different expressive intentions (we might speak of "efforts" using Laban's term), say, light, soft, heavy, nervous, or basic emotions like happy,sad... An archive of microdances has been created and is currently studied to identify relevant cues and find suitable computational models to extract such cues in real time. Humans (spectators) evaluate each microdance, in order to test how the intended expressiveness is perceived. We developed some computational models to analyse these expressive intentions and then we compare with spectators' rating of the same dance fragment. Implementations of our computational models on real-time analysis of expressive content (for use in real time apps including interactive dance/music and other artistic apps) are available as an external library (Motion Analysis) of our open software platform EyesWeb (www.eyesweb.org : you may download the full software and join the public newsgroup). This Motion Analysis Lib will be released publicly in the upcoming version 2.5 (released next May 7, 2002). We have a paper in preparation on some our recent work on computational models and implementations. Comments and feedback welcome. Best regards, Antonio Camurri
This archive was generated by hypermail 2b30 : 04/03/02