Reservoir A/V performance at Tedx KeioU conference
|
This study explores new possibilities for artistic expression by utilizing reservoir computing in an improvisational call-and-response session between a human performer (Professor Fujii) and a machine learning system.
By leveraging the dynamic internal state changes of reservoir computing, we developed a system capable of generating complex rhythms and a sense of groove, creating a structure where the performer and the system mutually influence each other. As a result, the reservoir computing model demonstrated its ability to provide new inspiration to the performer through improvisational responses, enabling co-creative performance. Furthermore, its integration with visual expression revealed new possibilities for artistic expression. Audio Visual Performance: Keigo Yoshida Drum Performance: Shinya Fujii Narration Vocal: Rinko Oka Audio Mixing: Makoto Nagase Audio Equipment Support: Onzo Kobo Broad Cast Support: CIMA Co., Ltd Production Support: Kenshiro Taira Production Management: TedxKeioU Reference Paper: Oscillations create groove: A reservoir model for learning complex drumming performances Reservoir Model Provision: Yuki Kawai, Minoru Asada from Symbiotic Intelligent Systems Research Center, Institute for Open and Transdisciplinary Research Initiatives, Osaka University, Suita, Osaka, Japan, Shinya Fujii from Faculty of Environment and Information Studies, Keio University, Fujisawa, Kanagawa, Japan supported by JST 戦略的創造研究推進事業（さきがけ） |