Human Motion — Generative Models
Understanding and Generating Human Motion
2019 — 2024
Through the EU Horizon 2020 EnTimeMent project, I worked on group behavior recognition, controllable 3D motion generation, and dance style transfer for interactive media and social robotics.
Human motion is shaped by physical structure, temporal context, and interaction with other people. I explored these connections by recognizing behavior in social groups, generating movement from multimodal signals, and transferring motion style while preserving its underlying content.
