Your homography expertise can directly improve the next generation of computer vision and AI systems. You’ll help train models to get spatial reasoning right, not just “recognize stuff,” so the outputs are accurate, consistent, and usable in real-world vision and graphics work.
About Invisible
Invisible supports AI development through high-quality training data and expert evaluation. They work with specialists to review, refine, and improve AI outputs so models perform better across complex, real-world problem spaces.
Schedule
Contract, freelance, remote (U.S. listed).
Hours vary based on your availability and project needs.
You provide a secure computer and stable internet connection.
What You’ll Do
⦁ Review and annotate AI-generated content related to homography and projective geometry
⦁ Evaluate model outputs for mathematical correctness, clarity, and multi-view consistency
⦁ Identify reproducible error patterns and document failure modes for model improvement
⦁ Select and label static points across images to support transformations and evaluation tasks
⦁ Conduct manual and synthetic homographies and validate transformation accuracy
⦁ Collaborate with the team to refine prompts, evaluation methods, and domain guidelines
⦁ Provide actionable feedback that improves reliability for real applications (robotics, AR, 3D reconstruction, mapping)
What You Need
⦁ Foundational knowledge of homography and projective geometry (coursework, research, or applied work)
⦁ Experience in computer vision, robotics, graphics, or related technical fields
⦁ Strong analytical thinking and the ability to spot subtle mathematical or reasoning errors
⦁ Clear written communication for documenting issues and giving structured feedback
⦁ High attention to detail and consistency when labeling or validating transformations
⦁ Degree in mathematics, computer science, engineering, or similar is a plus
Benefits
⦁ Flexible, remote contract work
⦁ Pay range of $8–$65/hour (rate based on experience, expertise, and location)
⦁ Work that directly influences model quality and real-world usefulness
⦁ Exposure to cutting-edge AI tools and evaluation workflows
⦁ Project-based opportunities that can scale with performance and availability
Roles like this move fast, and strong specialists get picked up early.
If you’re ready to turn homography skill into real model impact, get your application in and set a confident rate that matches your depth.
Happy Hunting,
~Two Chicks…