Who are we?
Our international team includes leading experts on the evaluation of gesture-generation models, as well as experienced model developers and research engineers with experience in deploying gesture-generation models. Together, we are developing the leaderboard and its associated tooling, and we will be responsible for managing and funding year-round crowdsourced human evaluations.
Our cumulative experience covers all major aspects of gesture-generation research, including:
- crowdsourced evaluation: e.g., organising the GENEA Challenges in 2020-2024 (the leading large-scale human evaluation efforts in gesture generation to date!)
- data collection: e.g., TED Gesture dataset
- model development: e.g., Gesticulator (ICMI 2020 Best paper), Gesture Generation from Trimodal Context (SIGGRAPH Asia 2020), StyleGestures (EUROGRAPHICS 2020 Honourable mention), Listen, Denoise, Action! (SIGGRAPH 2023), AQ-GT (ICMI 2023 Best paper)
- visualisation tooling: e.g., Blender, Maya, and Unreal Engine development
Organizers
Gustav Eje Henter
Assistant Professor
KTH Royal Institute of Technology
Head of Research motorica.ai, Sweden
Scientific advisors
While the organising team handles day-to-day operations, we are fortunate to be advised in key strategic decisions and the leaderboard methodology by three leading experts of nonverbal communication, visual perception, human-agent interaction, and motion capture:
Rachel McDonnell
Trinity College Dublin, Ireland
Michael Neff
University of California, Davis, USA
Stefan Kopp
Bielefeld University, Germany