Video Annotator

Position Overview

This role is a fully remote, full-time position focused on detailed video annotation tasks for a Vision-Language-Action dataset project. The main responsibility is to carefully review short video clips of physical activities and create precise records by breaking down each video into distinct steps paired with clear, natural language descriptions describing the actions shown.

Key Responsibilities

• Carefully view assigned videos in their entirety to understand the task and context before starting annotation work.
• Divide videos into discrete segments or steps based on clear visual action boundaries.
• Craft step-by-step descriptions for each segment that are concise, precise, grammatically correct, and reflective of the on-screen activities.
• Ensure all annotations focus on actions, avoiding ambiguity, and maintain consistent tense, tone, and phrasing throughout the dataset.
• Accurately note start and end timestamps for each video segment.
• Adhere strictly to annotation instructions, guidelines, category taxonomies, and style manuals as provided.
• Identify and report any videos that are corrupted, mislabeled, unclear, or unsuitable for annotation.
• Engage in calibration meetings and incorporate feedback from quality assurance teams to enhance annotation accuracy and uniformity.
• Fulfill set productivity and quality benchmarks on a daily and weekly basis without sacrificing annotation quality.
• Maintain strict confidentiality regarding all project materials and data.

Required Qualifications

• Advanced English proficiency at the C1 level or higher, including strong grammar, vocabulary, and sentence formulation skills.
• Exceptional attention to detail with the ability to detect subtle visual scene changes.
• Strong written communication capabilities to convey actions clearly and succinctly.
• Comfortable working with video annotation software and related tools; training will be provided.
• Capable of working independently while following detailed guidelines and sustaining consistency across repetitive tasks.
• Access to a reliable internet connection and a computer with a modern web browser.
• Basic digital skills including file management, use of web tools, and spreadsheets.

Preferred Qualifications

• Previous experience in annotation, data labeling, transcription, or content moderation roles.
• Knowledge of AI and machine learning concepts, especially related to computer vision, natural language processing, or robotics datasets.
• Experience creating instructional or procedural written content such as how-to guides, standard operating procedures, or recipes.
• Familiarity with annotation platforms such as CVAT, Label Studio, Scale, or similar tools.

About The Global Talent Company

The mission of The Global Talent Co. is to connect exceptional global talent with leading world brands, believing that access to opportunity should be universal and not limited by birthplace.