Research Scientist – Body Pose Detection

The Role

While our existing perception division handles state estimation and spatial mapping, this role is dedicated to one of the most critical bottlenecks in embodied AI: full-body kinematics, human locomotion, and human-scene interaction. We are hiring a Research Scientist to architect and train proprietary foundation models from scratch focused on 3D human body tracking and articulated pose estimation.

Your core mandate is twofold: building our in-house equivalents to cutting-edge 3D human body and mesh recovery architectures, and developing highly robust interaction models tailored for complex, real-world environments characterized by severe occlusions and dynamic motion. Beyond these core pillars, you will serve as a lead problem-solver for emergent perception challenges as our hardware and downstream robotics needs evolve.

To achieve this, we can provide a massive, continuous stream of high-quality, proprietary ground-truth human motion data captured by our infrastructure. You will use this data advantage to train networks that surpass current public baselines, owning the complete human-scene perception loop for our data engine.

What You'll Work On

Architecting Proprietary Articulation Models

• Zero-to-One Model Development: Design, implement, and train state-of-the-art networks for 3D human pose estimation, dense full-body mesh recovery, and kinematic tracking.

• Large-Scale Distributed Training: Scale multi-view and temporal ML architectures across multi-GPU clusters to handle massive, multi-modal datasets of humans navigating and interacting with their environments.

• Loss & Architecture Innovation: Push the boundaries of current paradigms by developing novel loss functions that enforce biomechanical constraints, temporal smoothness, postural balance, and physical plausibility.

Human-Scene Interaction (HSI) & Complex Motion Modeling

• Dynamic Scene Understanding: Build and train custom architectures capable of handling extreme motion blur, severe self-occlusion, and multi-person crowding inherent in real-world human behavior.

• Allocentric & Egocentric Tracking: Use your models to track human bodies through complex spaces, mapping foot-to-ground contact, joint torques, and environmental affordances to provide rich regularization for downstream action-conditioned robotics models (especially humanoid robots).

Emergent Perception R&D

• Rapid Prototyping: Tackle novel, unmapped AI challenges as they arise. You will rapidly prototype and deploy new models for tasks spanning fine-grained action segmentation, intent prediction, and novel hardware sensor integrations.

• Agile Problem Solving: Pivot to resolve sudden algorithmic bottlenecks in the data engine, adapting the latest research to unblock new product capabilities for our robotics customers.

Dense Contact & Physics-Aware Tracking

• Interaction Integration: Connect the outputs of your foundational tracking models into highly optimized pipelines that reason about physical contact surfaces, gravity, and momentum, directly bridging the gap between human video data and robotic control/locomotion policies.

What You Bring

Required Background

• Deep expertise in Deep Learning, 3D Computer Vision, and specifically Articulated Tracking / Human Body Pose Estimation.

• Proven experience training large-scale vision models from scratch, not just running inference or fine-tuning existing checkpoints.

• Strong theoretical and practical understanding of parametric human body models (e.g., SMPL, SMPL-X, GHUM, MHR, SOMA-X), inverse kinematics, and dense mesh estimation.

• Mastery of PyTorch and deep learning scaling frameworks.

• Experience handling and curating massive, multi-terabyte image and video datasets for training.

• Comfortable operating in a fast-paced environment where priorities can shift rapidly to capitalize on new research or hardware capabilities.

Strong Signals:

• First-author publications in top-tier venues (CVPR, ICCV, ECCV, NeurIPS) focusing on 3D human pose tracking, human-scene interaction (HSI), human motion capture, or human mesh recovery.

• Specific experience working with massive human motion and interaction datasets (e.g., AMASS, Human3.6M, EgoBody, PROX) and solving the unique optimization challenges they present.

Why This Role?

• The Data Advantage: You will have access to a scale and quality of proprietary spatial and temporal ground truth for human motion that most academic researchers only dream of.

• Pure R&D & Model Ownership: You are not maintaining legacy systems; you are given a blank slate and the compute resources to build the state-of-the-art.

• High Impact: The kinematic priors and interaction models you architect will directly define how the next generation of embodied AI agents—from mobile manipulators to humanoid robots—learn to physically navigate, balance, and interact with the world.

Warning: Research Scientist positions require hyper-specific expertise. Please limit your applications to one research role. Applying to multiple Research Scientist positions suggests a lack of focus and may result in the rejection of all submissions. You may, however, apply to other non-research roles alongside your research application.