"'This study is an exciting contribution for paleontology and an objective, data-driven way to classify dinosaur footprints - something that has stumped experts for over a century,' said Professor Steve Brusatte, an author of the study. 'It opens up exciting new possibilities for understanding how these incredible animals lived and moved, and when major groups like birds first evolved. 'This computer network might have identified the world's oldest birds, which I think is a fantastic and fruitful use for AI.'"
"To train the app, the researchers fed it nearly 2,000 fossil footprints alongside millions of variations to mimic changes such as compression and edge displacement. Amazingly, tests have revealed that DinoTracker can now identify dinosaur footprints with 90 per cent accuarcy - even for contentious species. One of the most interesting discoveries by the app was the resemblance between several dinosaur tracks and those left by birds."
An AI app called DinoTracker analyzes dinosaur footprints using advanced algorithms trained on nearly 2,000 fossil prints and millions of simulated variations to mimic compression and edge displacement. The app achieves about 90% identification accuracy, including for disputed specimens, reducing bias from manual methods. Several fossil tracks exhibit features shared with extinct and modern birds, implying either an avian origin tens of millions of years earlier than current estimates or convergent bird-like foot morphology in some dinosaurs. DinoTracker provides an objective, data-driven method for footprint classification and enables new investigation into prehistoric locomotion and evolution.
Read at Mail Online
Unable to calculate read time
Collection
[
|
...
]