Remote Video Annotation Expert: Action & Hand-Pose QA

OpenTrain AI; Northern (KY); Hybrid; USD 21,000 - 28,000; Part time

Benefits offered by this job

Remote work

Contractor role

Job summary

OpenTrain AI is seeking detail-oriented Video Annotation Experts to review, validate, and improve annotations across diverse video datasets. You will focus on accuracy, consistency, completeness, action annotations, and hand‑pose data.

This is a remote contractor role for candidates who are fluent in English and based in one of the eligible countries. No prior AI experience required, while knowledge of video labeling workflows and computer vision concepts is preferred.

Qualifications

• Fluent English and ability to work independently in a remote setting.
• Experience with video labeling workflows and datasets.
• Knowledge of action recognition and hand-pose analysis beneficial.

Responsibilities

• Review video annotations for accuracy, consistency, and completeness.
• Identify and tag quality issues or discrepancies in annotations.
• Make precise corrections to improve overall annotation quality.
• Evaluate English-language action annotations against guidelines.
• Assess and validate hand-pose data using quality standards.
• Provide concise written feedback and suggest guideline improvements.
• Document findings and corrections using standardized formats.
• Follow project instructions and meet deadlines.

Skills

Fluent English

Attention to detail

Video labeling workflows

Video understanding

Hand-pose analysis

Tools

Video labeling tools

Job description

OpenTrain AI is seeking detail-oriented Video Annotation Experts to review, validate, and improve annotations across diverse video datasets. You will focus on accuracy, consistency, completeness, action annotations, and hand‑pose data.

This is a remote contractor role for candidates who are fluent in English and based in one of the eligible countries. No prior AI experience required, while knowledge of video labeling workflows and computer vision concepts is preferred.