Computer Vision ML / Research Engineer

Description

The company is an industrial computer vision team building a perception system that converts stereo camera images into accurate, auditable inventory measurements.

You will own the core vision and machine learning pipeline and help move research methods into dependable production use.

Responsibilities

• Develop and refine stereo reconstruction, depth estimation, segmentation, and 3D volume measurement workflows.
• Integrate computer vision models and build point-cloud cleanup, ground-plane estimation, and scene-stitching methods.
• Assess reconstruction quality and volume accuracy, and dig into failure modes using real-world sensor data.
• Set calibration assumptions and create confidence metrics for measurements that must hold up in production.
• Turn research pipelines into robust systems for industrial inventory measurement.

Requirements

• You have four or more years of experience in computer vision or perception machine learning engineering.
• You are strong in Python and PyTorch and have used OpenCV and Open3D.
• You have hands-on work with stereo vision, depth estimation, point clouds, 3D reconstruction, and 3D geometry.
• You have shipped computer vision systems on real-world sensor data, including camera calibration and volume reconstruction.
• A background in robotics perception, autonomous systems, industrial inspection, photogrammetry, or similar fields is valued.
• Experience with CUDA, SLAM, YOLO-style models, FoundationStereo, or GPU cloud computing is a plus.
• A relevant master's or doctoral degree is beneficial.

Conditions

Base salary is $180,000 to $245,000 per year.

Visa sponsorship is available.

The role is on-site in San Francisco, California, United States.