We train models to understand the world from the sky
Aerial imagery has never been more prevalent or more valuable than it is today. We train, test, and tune the models that use those satellite images to understand the world.
Whether you need bounding boxes, pixel labeling, or just general image categorization, Figure Eight can help. We have a proven computer vision solution that can work with whatever ontology of classes you need to make your models a success.
We help build the annotated images that power the most accurate models in the space. We also test and tune these models by validating (or correcting) less confident outputs so they get better through the iterative process of active learning.
Our machine learning team has decades of computer vision experience. They can help you with everything from prioritizing the right kinds of data, tuning your algorithms, or looking at your approach holistically and making the suggestions that will elevate your model’s success.
Director of U.S. Search
Executive Vice President, Measurement and Insights
Senior Research Lead
Before we can think critically about computer vision, we need to take a moment to appreciate our own human vision...Read More