Top Computer Vision Jobs in Seattle, WA

Design, develop, and deploy robust perception systems for earthmoving autonomy: sensor fusion, scene understanding, obstacle detection, depth and terrain estimation, SLAM/localization, and optimized real-time inference on edge hardware. Build production ML pipelines, integrate perception with autonomy stacks, validate with field data, and collaborate across robotics, controls, and systems teams.

Develop and extend a C++ annotation and reporting tool for computer vision and robotics. Implement tracking, detection, deeper annotations, quality measurement, automated reporting, and automated tests. Work with OpenCV, SQL/reporting services, HTML/JavaScript/CSS, MSBuild, and Linux while collaborating with applied scientists and engineers.

Design and implement high-fidelity spatial mapping and real-time object recognition pipelines; integrate vision outputs with robotic manipulation and navigation systems; and train and fine-tune Vision-Language-Action models to improve robotic autonomy and perception.

Own computer-vision accuracy end-to-end: measure per-customer metrics, investigate misclassifications, curate and label datasets, prioritize retraining, and build continuous learning pipelines. Partner with ML/CV engineers and stakeholders to define acceptance criteria, translate findings into actionable fixes, and eventually implement tooling and retraining to ship regular model improvements.

Design real-world computer vision tasks, build reference solutions and evaluation components in Python, assess AI model outputs, identify failure modes, and develop structured rubrics and guidelines. Collaborate with experts to maintain high-quality training data and evaluation standards for multimodal and generative vision systems. Part-time, remote contractor role (~20 hours/week).

Senior product engineer owning end-to-end product surfaces for a computer vision platform. Ship features from idea to production, prioritize roadmap, design UX details, integrate APIs, improve annotation and model-assisted labeling, and collaborate with engineering and users to drive product impact.

Design, improve, and deploy production-grade computer vision systems for sports video: detection, multi-object tracking, camera calibration, homography, field registration, identity association, and robustness across varied cameras and video quality. Lead experiments, adapt models to new domains, collaborate with data and engineering teams, and drive end-to-end solutions from prototype to production.

Design, train, and deploy production computer vision and vision-language models for retail product recognition. Optimize models for edge devices, build data pipelines and annotation workflows, fine-tune open-source VLMs, develop VLA pipelines, mentor engineers, and drive CV infrastructure and MLOps best practices.

Develop end-to-end perception solutions for autonomous driving, leveraging deep learning and data-driven approaches to enhance system accuracy and efficiency.

Leads the technical direction of an autonomous vehicle perception stack. Responsibilities include architecting scalable perception systems, developing scene detection and runtime activation, improving robustness against sensor degradation and adverse weather, integrating multimodal models, managing ML and data pipelines, monitoring field performance, leading testing and feature delivery, and mentoring engineers. Requires strong robotics, computer vision, machine learning, real-time systems, software architecture, and team leadership experience.