Computer Vision Data Annotator
Job Description
The Computer Vision Data Annotator is responsible for preparing and annotating visual datasets used to train, validate, and evaluate computer vision and machine learning models. The role involves identifying visual objects, classifying image content, annotating spatial relationships, and creating structured visual information that enables AI systems to recognise and interpret objects, scenes, and activities accurately.
The specialist will work closely with computer vision engineers, machine learning researchers, and data scientists to develop high-quality annotated datasets for applications such as autonomous vehicles, robotics, industrial automation, medical imaging, facial analysis, and intelligent surveillance systems. The position requires a strong understanding of visual data annotation techniques, object recognition principles, and dataset quality management.
Key Responsibilities:
Prepare and annotate image and video datasets for computer vision applications.
Identify and classify objects, scenes, and visual patterns.
Apply bounding boxes, polygons, and semantic segmentation masks.
Annotate spatial relationships between objects within images.
Identify object attributes, orientations, and visual characteristics.
Support image classification and object recognition projects.
Review datasets for annotation errors and inconsistencies.
Identify difficult visual cases and escalate complex annotation issues.
Collaborate with engineering teams to improve annotation accuracy.
Maintain structured records of completed computer vision annotation tasks.
Qualifications and Requirements:
Bachelor’s degree in Computer Science, Artificial Intelligence, Data Science, or a related discipline.
Practical experience in image or video annotation.
Understanding of computer vision and machine learning concepts.
Familiarity with CVAT, Labelbox, Supervisely, or similar platforms.
Knowledge of object detection and image segmentation techniques.
Strong analytical and visual interpretation abilities.
