Son O.

Key Skills

Software

Other

Top Subject Matter

Computer vision (autonomous driving, retail shelf analysis, medical imaging, surveillance/activity recognition)

Computer vision research and crowdsourced AI labeling projects (pose estimation, instance segmentation, scene understanding)

Legal Services & Contract Review

Top Data Types

Video

Text

Document

Top Task Types

Tracking

Segmentation

Freelancer Overview

Data Labeling Specialist (Image & Video) — OpenTrain.AI (Remote, Lagos). Core strengths include OpenTrain.AI and Other. Education includes Bachelor of Science, University. AI-training focus includes data types such as Video and Image and labeling workflows including Tracking and Segmentation.

Labeling Experience

Data Labeling Specialist (Image & Video) — OpenTrain.AI (Remote, Lagos)

VideoTracking

Contributed to AI/ML training datasets by annotating over 50,000+ image frames and performing video sequence labeling with object tracking across frames. Followed client ontologies and edge-case guidelines while maintaining temporal consistency for surveillance and activity-recognition use cases. Validated work through internal QA audits and inter-annotator agreement to sustain high annotation accuracy and reduce label variance. • Bounding boxes, polygons, and semantic masks for multi-class object detection • Video object tracking, frame-to-frame temporal consistency, and interpolation • Annotation calibration sessions to align team labeling standards • Escalation of ambiguous or out-of-scope samples to refine ontologies and improve model performance

2023 - Present

Freelance Image Annotation Contributor — Independent AI Projects (Lagos)

OtherImageSegmentation

Performed AI annotation work for independent/crowdsourced research projects, delivering polygon segmentation and keypoint labeling for pose estimation and instance-segmentation tasks. Completed quality-reviewed batches of image classification and scene labeling with strong accuracy and throughput. Managed task queues and deadlines across multiple concurrent annotation assignments while meeting delivery targets with minimal QA rejections. • Polygon segmentation for instance segmentation research datasets • Keypoint/pose annotation for pose-estimation research tasks • Image classification and scene labeling for research workflows • Self-managed task queues to meet deadlines and quality requirements

2022 - 2023

Education

Work History