Research Scientist – Full Body Motion

The role

You will build a model to replace current pipelines by generating full-body pose from partial egocentric evidence. This involves training generative motion models and implementing physics-based tracking or contact-aware optimization.

What you will do

• Train generative motion models like diffusion, autoregressive or masked on mocap-scale and rig data.
• Implement physics-based tracking or imitation in simulation or contact-aware optimization.
• Deliver per-frame foot and body contact and the ground plane with calibrated confidence.
• Develop MHR / SOMA-X native output including pose priors and constraints.
• Define lower-body, feet and contact metrics and commission captures.
• Prototype and deliver runnable models to production owners.
• Publish research findings.

Required qualifications

• PhD or equivalent body of work in physics-based character animation, human motion generation, or 3D human pose and shape.
• Publications at SIGGRAPH / SIGGRAPH Asia, CVPR / ICCV / ECCV, NeurIPS / ICLR or CoRL.
• Experience training diffusion, autoregressive or masked motion models on mocap-scale data conditioned on constraints.
• Experience with motion imitation and tracking for simulated humanoids in the DeepMimic / AMP / MaskedMimic lineage.
• Comfortable in Isaac Gym / Isaac Lab or MuJoCo with reinforcement learning at scale.
• Knowledge of SMPL-X, MHR or SOMA-X, retargeting, and contact and collision modelling.
• Clean PyTorch research code.
• Ability to work on noisy, real, large-scale data.

Nice to have

• Egocentric or partial-observation full-body estimation
• Human–scene interaction and contact: scene-aware motion synthesis, volumetric body models
• Motion cleanup or denoising of estimated or corrupted motion
• Reusable motion priors for control, generative controllers, or large-scale motion tokenisation
• Having shipped a research model into a product or a game engine
• Familiarity with Momentum / MHR tooling

About the company

Mecka AI is building the data infrastructure layer for robotics and embodied AI. They design and operate global systems for data capture, labeling, and hardware-enabled workflows for AI labs and robotics companies.