Senior Computer Vision Engineer, Egocentric

This is a fully remote position, open to applicants in United States.

📋 Description

• Define and execute the egocentric video data product across various quality levels, ranging from RGB video to depth-enhanced and multimodal capture with hand pose and annotations.

• Establish product priorities driven by buyer demand while upholding dataset quality standards.

• Initiate and manage warehouse capture operations, including the selection of camera and rig hardware, enrollment, edge processing, and dataset delivery pipelines.

• Collaborate with warehouse operations, engineering teams, and clients to ensure datasets are delivered according to specifications and timelines.

• Develop perception systems for tasks such as detection, tracking, segmentation, depth/3D reconstruction, 6DoF, and multi-view 3D hand/body pose estimation.

• Implement VLM-assisted and automated labeling processes with human-in-the-loop quality assurance, while integrating annotation tools.

• Oversee camera calibration, epipolar/multi-view geometry, frame-accurate synchronization, and triangulation for multi-camera and egocentric rigs.

• Design, refine, optimize, and deploy computer-vision and multimodal models on extensive unstructured video datasets.

• Ensure models are reproducible and ready for production.

⛳️ Requirements

• Proven experience in establishing and scaling an egocentric perception stack within a robotics or AI data organization.

• Demonstrated capability in building and managing a similar product from inception to completion.

• Over 8 years of experience in developing and deploying production computer-vision/perception systems, or hold an MS/PhD in CV, ML, or robotics, supplemented by 6+ years of practical experience.

• Strong expertise in computer vision and tooling, including the design, training, and debugging of CNNs and vision transformers from the ground up.

• In-depth knowledge of camera calibration, depth estimation, and 2D/3D pose estimation.

• Comprehensive ownership of perception challenges, encompassing data and model design through to evaluation, optimization, and deployment.

• Proven ability to set the technical direction for a team or significant workstream.

• Capability to navigate ambiguous 0→1 challenges to create functional systems with limited resources.

• Experience working with large unstructured video/multimodal datasets and accuracy evaluation tools.

• Proficient in Python with a solid foundation in software engineering principles.

• C++ proficiency where performance is critical.

• Must possess authorization to work in the U.S.

🏝️ Benefits

• A structural advantage that is difficult for startups to replicate.

• Direct access to the rapidly expanding buyer market in AI.

• Opportunity to work directly alongside the CTO/Co-Founder as your partner.