About the postition
As an AI Research Scientist at MVNT, you will pioneer the world's first dance-specific AI generation models, driving research in motion synthesis and cultural representation learning. You will collaborate with our team to build breakthrough dance AI that powers mvntSTUDIO and transforms how the gaming industry creates authentic movement. This role is perfect for research-minded scientists who want to build state-of-the-art dance and motion AI systems.
In this role, you will:
Develop cutting-edge dance motion generation models using our proprietary 3D mocap and 2D video datasets
Research dance synthesis that understands stylistic and spatial-temporal nuances across dance
Optimize training and inference performance of our core dance generation model stack
Drive AI research from ideation through prototyping to production integration in mvntSTUDIO
Design and maintain data preprocessing pipelines for 3D motion capture and 2D pose estimation datasets
We'll be perfect if you have:
Currently enrolled in or completed at least one year of master's/doctoral studies in AI, ML, Computer Vision, or related fields
Publications in top-tier AI conferences (NeurIPS, ICML, ICLR, AAAI, CVPR, ICCV, ECCV, SIGGRAPH)
Proficiency in Python and PyTorch
Research experience in Representation Learning, Diffusion Models, Action Recognition, Video Understanding, or Human Motion Analysis
Genuine affinity with dance and cultural understanding of movement (essential for our mission)