← Back to Projects
3D Motion Synthesis

3D Motion Synthesis

2024

Music-Generated Context-Aware 3D Movements

Transformers, Diffusion, PyTorch, 3D Movements

AI/ML

Integrating Transformer-Based AI-Generated Dance Movements Into HRP-4 for Human-Robot Artistic Collaboration

Hui-Ting Hong, Chu-Yin Chen, and Abderrahmane Kheddar

International Workshop on Human-Friendly Robotics (HFR) 2024

DOIPDF

This paper presents the integration of dance movements generated from a transformer-based diffusion model onto the HRP-4 humanoid robot using the open-source framework mc_rtc. We explore how AI-generated movements can facilitate meaningful collaboration between robots and artists and discuss the methodology for incorporating these movements into the robot’s control framework. Furthermore, we examine the artistic perspectives and user engagement that arise from this integration.

3D Motion Synthesis — 1
3D Motion Synthesis — 2

Team & Partners

  • Labo. INREV-AIAC, Paris 8
  • LIRMM — Laboratoire d’informatique, de robotique et de microélectronique de Montpellier