Research Robotics/Computer Vision Engineer
Core
Developing perceptive, intelligent, and adaptable robotic systems focused on 3D computer vision and autonomous navigation.
Role type
Research Robotics/Computer Vision Engineer
Builds
Perception pipelines, SLAM systems, learning-based algorithms for robotic control, and data infrastructure for robot manipulation training.
Domain
Robotics, Computer Vision, Autonomous Systems
Deliverable
production ML models | infrastructure
Required skills
3D scene reconstruction (monocular videos, meshes, point clouds, Neural Radiance Fields, Gaussian Splats), rigid and articulated object reconstruction from videos, generative computer vision (diffusion models), attention-based models for perception, Neural Architectural Search (NAS), cloud-based distributed training (AWS, Google Cloud, Vertex AI)
Preferred skills
null
Technologies
PyTorch, AWS, Google Cloud, Vertex AI
Responsibilities
Implement perception for safe exploration and navigation; reconstruct 3D scenes using monocular images and optimize 3D SLAM; develop software tools for visual-only robot localization; build software for life-long mapping via pose-graphs; perform visual servoing for robot motion control; research techniques to handle glare during mapping; build data pipelines for streaming hand movements to train manipulation tasks; maintain camera and 2D lidar navigation stacks.
Seniority
Mid-level, hands-on IC
