An AI reinterprets the ephemeral nature of human dance using trails of laser light projected on the surface of phosphorescent rocks
A collaboration with artist Eugenia Kim exploring the meaning of embodiment as seen by a machine. Dance and movement data that has been collected via methods such as mocap systems and processed with machine learning algorithms. In existing movement visualisation works, the human form is still emphasized even when heavily abstracted, and the data remains in a digital and/or virtual realm. In response to this we explore the application of laser projections of motion trails to bring data into physical reality, thus metaphorically giving a “body” to generated movement. Somatic movement improvisations (i.e. Contact Improvisation and the Skinner Releasing Technique) will be used for training to teach the attributes of human movement rather than the vocabulary of a set dance technique.
The project began as a philosophical conjecture on how a machine might visualise its own body based on its understanding of human somatic movement practises. Lasers were considered to be the most appropriate method for translating from digital data into the physical realm as it is simultaneously tangible and malleable. A proof-of-concept to test the interplay between movement and light was then developed using motion capture data of the choreographic work Oncheon. The mocap data held special significance for two reasons. First, it combined the structure of salpuri, a traditional Korean shamanic dance to expel spirits, with contemporary dance and somatic movement. Second, Oncheon was performed and recorded despite triggering severe chronic pain and fatigue. It is unlikely that it can ever be performed again by the original choreographer/performer.
Publications
- Kim, E. S., Haebich, J. K. & Cassinelli, A., 2022, Radiant Soma, ACM SIGGRAPH 2022 (Special Interest Group on Computer Graphics and Interactive Techniques) Vancouver, Canada. PDF
- Kim, E. S., Haebich, J. K. & Cassinelli, A., 20 Apr 2022, Radiant Soma: Visualization of Movement Through Motion Capture and Lasers, (Accepted/In press/Filed) MOCO ’22: Proceedings of the 8th International Conference on Movement and Computing: International Conference Proceedings Series. Association for Computing Machinery. PDF
- Kim, E.S., Sandor, C., Haebich, J. and Cassinelli, A., 2021, June. Playing with Soma: Speculating on the Physical Body and Somatic Practice of AI. In Art Machines 2: International Symposium on Machine Learning and Art 2021 (AM2) (pp. 31-38). School of Creative Media, City University of Hong Kong. PDF