An advanced AI tool known as DinoTracker is revolutionizing the study of dinosaur footprints by analyzing fossilized tracks to identify the species that may have made them. Developed by researchers based in Edinburgh and Berlin, the model was trained using a dataset comprising 2,000 authentic fossil footprints alongside millions of simulated examples. This sophisticated system examines various attributes of the footprints, such as toe spacing, heel placement, the area in contact with the ground, and how weight is distributed across the foot. Remarkably, the tool's predictions align with the assessments of paleontology experts in 90% of instances, even when dealing with species that are difficult or contentious to classify. DinoTracker has already proven invaluable in deciphering enigmatic footprints that date back 170 million years, discovered on the Isle of Skye in Scotland. These tracks are now believed to belong to some of the earliest ancestors of duck-billed dinosaurs. The DinoTracker software is available for download on GitHub.
Informational material. 18+.
