DATA Annotation Specialist

DATA ANNOTATION SPECIALIST

Image & Video Labelling • Computer Vision • Quality Assurance

Role Overview

We are seeking a detail-oriented Data Annotation Specialist to create high-quality labelled datasets for computer vision and machine learning applications. The role involves accurately annotating images and video frames using bounding boxes, polygons, segmentation masks, keypoints, and classification labels in CVAT or similar annotation platforms.

It requires precision, consistency, guideline discipline, quality awareness, and clear communication in English.

Key Responsibilities

• Label and annotate image and video datasets according to project-specific guidelines and class definitions.
• Create and refine 2D bounding boxes around objects of interest with accurate placement and minimal excess background.
• Draw polygons and segmentation masks for irregular objects, lanes, drivable areas, obstacles, and other defined regions.
• Apply object classifications, attributes, tags, and keypoints where required.
• Use CVAT or comparable annotation tools to create, edit, review, and export annotations.
• Maintain consistency across frames, object classes, attribute values, and annotation sequences.
• Identify unclear, ambiguous, occluded, truncated, or difficult cases and escalate them for clarification.
• Perform self-quality checks and peer reviews against annotation guidelines and acceptance criteria.
• Correct rejected annotations and incorporate reviewer feedback within agreed timelines.
• Track assigned tasks, completed frames, quality issues, and rework status accurately.
• Protect confidential data and follow applicable information-security and data-handling requirements.

Required Qualifications & Experience

• Hands-on experience in data labelling or annotation for images, videos, or computer vision datasets.
• Practical experience with bounding-box annotation and polygon-based annotation.
• Experience using CVAT; experience with Labelbox, Supervisely, Label Studio, V7, VIA, or similar tools is also relevant.
• Good visual attention, spatial judgement, and ability to identify small or partially visible objects.
• Ability to understand and consistently apply detailed annotation guidelines.
• English fluency, with the ability to read instructions, document issues, and communicate clearly with reviewers and project teams.
• Basic computer proficiency and confidence working with web-based applications, folders, and structured datasets.
• Ability to perform repetitive, accuracy-focused work while meeting productivity and quality targets.

Preferred Skills

• Experience with semantic segmentation, instance segmentation, keypoint annotation, lane marking, or object tracking.
• Exposure to automotive, ADAS, traffic-scene, manufacturing, or other computer vision datasets.
• Understanding of common annotation formats such as JSON, XML, CSV, COCO, Pascal VOC, or YOLO.
• Knowledge of quality-control workflows, inter-annotator consistency, and defect categorization.
• Basic understanding of machine learning or computer vision concepts is an advantage.

Education

Diploma in Computer Science/Information Technology or similar, bachelors degree in Science (BSc Computer Science/Information Technology or similar), ITI in Computer Operator and Programming Assistant/Information System or equivalent qualification.

Relevant experience in annotation, data operations, quality inspection

Performance Expectations

• Consistently meet defined annotation quality and acceptance criteria.
• Complete assigned work within agreed productivity and delivery targets.
• Minimize avoidable rework through careful self-checking.
• Raise questions early when guidelines or examples are unclear.