Publication

Human-AI Co-Dancing: Evolving Cultural Heritage through Collaborative Choreography with Generative Virtual Characters

Copyright

Cyber Subin

Lee Chia Yeh

Under review at ACM MOCO

Abstract

This paper presents a novel approach of "human-AI co-dancing" that combines human dancers with virtual dance partners powered by algorithms derived from traditional choreographic knowledge. Specifically, the principles of Thai classical dance "Mae Bot Yai" were translated into computational procedures capable of generating and transforming movements in real-time. A system was developed to allow dancers to explore these traditional vocabularies and principles generatively by improvising alongside the virtual agent. Multiple interaction modalities were devised, including direct voice control, mediated control by a secondary dancer/choreographer, and even audience participation through real-time voting of parameter changes. Preliminary co-creation rehearsals exhibited intriguing artistic results as hybrid movement aesthetics emerged from the synergy and friction between human and machine. This research illustrates promising potentials at the intersection of intangible cultural heritage, intelligent technology, and posthuman choreography, using computation to evolve traditional knowledge into novel forms for the future. 

Related Content