Remote Imagery Annotation Specialist for AI Training Data

Crossing Hurdles; United States; Remote; USD 28,000 - 83,000; Part time

Job summary

Crossing Hurdles is seeking an Imagery Analyst to support AI/ML training data projects. This remote contractor role offers part-time or full-time hours and an hourly compensation model.

You will review machine-generated bounding boxes, annotate objects in drone and maritime imagery, and ensure data labeling adheres to strict guidelines within a secure Navy environment. The role anticipates high throughput and the possibility to transition to video-frame annotation as the project grows.

Qualifications

• Demonstrated attention to detail and consistency in repetitive tasks.
• Experience with data labeling and bounding box annotation.
• Familiarity with aerial, maritime, or sensor imagery, daylight or thermal formats.
• Track record in high-volume, quality-controlled annotation environments.
• Strong written and verbal communication for reporting labeling requirements.

Responsibilities

• Review and refine machine-generated bounding boxes for drone and maritime imagery.
• Identify and annotate objects missed by AI models.
• Label data with consistent guidelines.
• Deliver annotated images within a secure Navy environment.
• Maintain steady throughput; compensation is based on image submissions.
• Transition to video-frame annotation as the project expands.
• Document ambiguous cases and clarify labeling guidelines.

Skills

Attention to detail

Data labeling

Bounding box annotation

High-volume annotation

Communication skills

Tools

Labeling software

Job description

Crossing Hurdles is seeking an Imagery Analyst to support AI/ML training data projects. This remote contractor role offers part-time or full-time hours and an hourly compensation model.

You will review machine-generated bounding boxes, annotate objects in drone and maritime imagery, and ensure data labeling adheres to strict guidelines within a secure Navy environment. The role anticipates high throughput and the possibility to transition to video-frame annotation as the project grows.