Top Computer Vision Jobs in Boston, MA

Leads the technical vision, research strategy, engineering execution, and organizational growth of a global computer vision platform. Oversees research and engineering teams, defines architecture and roadmaps, manages end-to-end MLOps across edge and cloud environments, establishes performance metrics, drives innovation, and mitigates technical risks. The role requires deep computer vision and machine learning expertise, executive-level technical leadership, large-scale organizational management, and experience deploying enterprise platforms in production.

Develop and own computer vision and deep learning systems for real-time crop-spraying applications. Build models for segmentation, classification, optical flow, and 3D reconstruction; optimize them for edge devices; and establish production MLOps, testing, and deployment practices. Collaborate with hardware, firmware, software, and product teams, analyze field performance, manage data pipelines, and continuously improve model accuracy, reliability, and inference speed.

Own end-to-end machine learning for robotic grasping and perception, including problem formulation, large-scale dataset generation and curation, model training, experimentation, evaluation, and verification. Develop vision models using synthetic data, self-supervised labeling, and active learning. Define metrics tied to robot performance, produce deployable model artifacts, conduct code reviews, mentor engineers, and contribute to the perception and grasping roadmap while partnering on runtime integration.

Build and deploy computer-vision perception systems for robot manipulation. Responsibilities include detection, segmentation, tracking, pose estimation, 3D scene understanding, camera calibration, dataset creation, evaluation, model selection, and diagnosing failures on physical robots. The role requires close collaboration across research, software, mechanical, and electrical engineering teams, plus maintaining reliable, measurable, and diagnosable perception systems.

Develop and prototype computer vision and machine learning solutions for robotic trailer unloading. Train and adapt models for perception tasks such as object detection, segmentation, grasp estimation, depth estimation, and anomaly detection. Troubleshoot real robotic systems, evaluate emerging robotics advancements, transition prototypes to product teams, communicate technical priorities, mentor junior engineers, and provide technical leadership on key projects.

Design and optimize computer-vision and robotics ML models for constrained edge devices, build large-scale training and fine-tuning pipelines, compress and deploy models to embedded accelerators, profile on-device inference for latency and accuracy, and write maintainable Python/C++ code while staying current on state-of-the-art techniques.

Develop and train computer vision and visual representation pipelines for robotic manipulation (RGB, RGB-D, depth, segmentation, pose, keypoints). Build perception models to support RL and imitation policies, improve synthetic/real data pipelines, evaluate models on real robots, calibrate sensors, and collaborate with robotics and simulation engineers to iterate on models and data for robust deployment.

Researches, designs, develops, trains, and deploys computer vision and machine learning models. Builds and maintains ML software, testing and validation processes, and documentation. Applies image and video processing, deep learning, and computer vision techniques across edge and cloud-based platforms while improving existing applications and performance.

Perform image segmentation and object labeling for computer vision datasets using polygons, bounding boxes, and semantic segmentation. Follow annotation guidelines, complete quality checks, meet daily productivity targets, communicate unclear tasks to team leads, and suggest workflow improvements. Work remotely with datasets involving roads, vehicles, satellite imagery, blueprints, roofing, and related visual content.

Review and annotate robotic-system video footage by identifying objects, movements, events, visual features, and anomalies. Provide feedback to improve video datasets and annotation consistency, collaborate with AI teams on annotation protocols, document observations, participate in quality reviews, and recommend improvements to video capture and labeling strategies.