Remote QC Reviewer - Human Pose Keypoint Annotations

Welo Global; United States; Remote; USD 8,000 - 11,000; Full time

Benefits offered by this job

Fully remote

Flexible schedule

Job summary

Welo Data is seeking a detail-oriented QC Reviewer to ensure the accuracy of human pose keypoint annotations on thermal images within Label Studio. You will review submissions, verify per-person counts, and provide actionable feedback to drive rework.

This remote, project-based role emphasizes careful evaluation of inferred keypoints and confidence indicators, with calibration sessions and ongoing documentation of review outcomes.

Qualifications

• Prior experience in computer vision annotation QA or related QC role.
• Strong understanding of human anatomy and pose structure.
• Exceptional attention to detail to catch subtle errors.
• Comfort with thermal/infrared imagery and its limitations.
• Ability to apply detailed annotation/QC guidelines consistently.
• Strong written communication for feedback to annotators.
• Familiarity with Label Studio review/approval workflows.
• Reliable, self-directed work habits meeting turnaround expectations.

Responsibilities

• Review annotator submissions against source thermal images for accuracy.
• Verify per-person keypoints and counts.
• Assess placement accuracy near anatomical centers.
• Evaluate plausibility of inferred/ooccluded keypoints.
• Check confidence flags and flag misused indicators.
• Identify systematic annotation errors for calibration.
• Provide actionable feedback within Label Studio review workflow.
• Document review outcomes and performance trends.

Skills

QA experience

Anatomy knowledge

Label Studio

Attention to detail

Thermal imagery

Communication

Tools

Label Studio

Job description

Welo Data is seeking a detail-oriented QC Reviewer to ensure the accuracy of human pose keypoint annotations on thermal images within Label Studio. You will review submissions, verify per-person counts, and provide actionable feedback to drive rework.

This remote, project-based role emphasizes careful evaluation of inferred keypoints and confidence indicators, with calibration sessions and ongoing documentation of review outcomes.