Video Annotator

About the Position

This fully remote, full-time role involves annotating videos for a Vision-Language-Action dataset project. You'll analyze short clips showing physical activities and segment them into meaningful steps, then draft precise, natural language descriptions for each segment to clearly convey the actions observed.

Key Duties

• Watch each video entirely before annotation to fully comprehend the task and context.
• Divide videos into logical, discrete steps based on visible action changes.
• Compose clear, concise, and grammatically accurate step-by-step descriptions that faithfully represent the on-screen actions.
• Maintain consistent tense, tone, and phrasing with an emphasis on action clarity.
• Mark precise start and end times for every segment.
• Adhere strictly to provided annotation guidelines, taxonomies, and style manuals.
• Report any videos that are unclear, corrupted, mislabeled, or otherwise unsuitable for use.
• Contribute to calibration meetings and apply feedback to enhance annotation precision and uniformity.
• Meet productivity and quality goals regularly without sacrificing accuracy.
• Uphold confidentiality of all project materials and data.

Required Qualifications

• English fluency at C1 CEFR level or higher, with a strong grasp of grammar and vocabulary.
• Exceptional attention to detail with the ability to detect subtle visual changes.
• Effective written communication skills to succinctly and accurately describe actions.
• Comfortable using video annotation tools and software (training provided).
• Self-motivated with the capacity to follow detailed instructions consistently through repetitive work.
• Access to a reliable internet connection and a computer equipped with a modern web browser.
• Basic computer skills including file management, use of web applications, and spreadsheets.

Preferred Qualifications

• Previous experience in data annotation, labeling, transcription, or content moderation.
• Knowledge of AI/ML concepts related to computer vision, natural language processing, or robotics datasets.
• Experience creating instructional or procedural documentation such as guides or standard operating procedures.
• Familiarity with annotation platforms like CVAT, Label Studio, or Scale.

Additional Information

The role supports the vision of bridging global talent with leading brands, promoting equal opportunity worldwide.