Dino Traces Found in Jurassic Park Analyzed With Artificial Intelligence

Artificial Intelligence Decodes Dino Traces Found in Jurassic Park
Artificial Intelligence Decodes Dino Traces Found in Jurassic Park

Artificial intelligence has been used by Australian researchers to uncover the truth behind 93-million-year-old dinosaur tracks.

Australia, part of the former supercontinent Gondwana, was still attached to Antarctica 93 million years ago, dinosaurs were still roaming the Earth, and the northeastern part of the Land Down Under was experiencing a very minor Cretaceous drama. Known as the Lark Quarry Dinosaur Tracks, the extraordinary paleontological find of 3.300 well-preserved dinosaur tracks was made near the city of Winton, in present-day Queensland, Australia. This place is now referred to as Australia's "Jurassic Park".

For years, scientists have speculated about the significance of this unique discovery.

Was this evidence of a rare dinosaur stampede escaping a theropod like T. rex, or was this just a well-traveled creek crossing?

Complicating this issue further is the fact that the tridactyl (three-toed) footprints of plant-eating ornithischians and theropod carnivores may be strikingly similar.

Anthony Romilio, a paleontologist at the University of Queensland, aimed to shed some light on the subject by using artificial intelligence to assist researchers in deciphering these prehistoric traces.

With the help of artificial intelligence, Romilio and his team succeeded in proving that the footprints previously thought to be theropods actually belonged to the more docile, plant-eating ornithischians. The researchers published their findings last week in the journal Royal Society Interface.

Romilio and colleagues created an artificial intelligence program called Deep Convolutional Neural Networks, which uses artificial neural networks to evaluate large areas of data using machine learning to make the technology big time-ready for footprint tracking.

Theropods and ornithischians, two dinosaurs linked to a 93-million-year-old mystery, have been fed more than 1.500 dinosaur footprints by researchers. Jens Lallensack, one of the study's authors, told The Royal Society that artificial intelligence "could potentially find ways to separate these categories that people haven't thought of yet," after combining features to help select categories.

Later, both human and artificial intelligence researchers took 36 additional tracks. According to Cosmos, AI detected footprints with 90% accuracy, while Romilio was only able to recognize tracks with up to 75% accuracy. Confident of the AI's precision, they let it examine the unidentified footprints in the Lark Quarry.

Conclusion? The Ornithischians, a group of dinosaurs from the Jurassic Period, left their mark. According to the University of California Museum of Paleontology, this group includes creatures such as duck-billed hadrosaurs, horned Ceratopsia, and armored Stegosauria.

This study is the final stage in Romilio's 12-year attempt to re-enact the drama in Winton, and marks the first time artificial intelligence has been used to analyze dinosaur tracks. It's a daunting task, given that the Lark Quarry Dinosaur Trails are also known as the Dinosaur Confluence National Monument.

In 2010, Romilio began exploring the idea that these marks were not caused by a confluence but rather by a well-used stream crossing. Some of Romilio's colleagues at the time were furious at the idea and referred to Romilio as an "iconoclast" who falsified his statistics. Since then, Romilio has penned the most important work on processing dinosaur tracks, and teaching AI to recognize footprints could be the next critical step in solving these Cretaceous conundrums.

According to Lallensack's statement to The Royal Society, "a scar is the product of many factors, including the anatomy of the foot, the density of the substrate, the movements of the animal that created the trace, and the changes that occur after the trace is created." Neural networks can truly transform the game to move forward by replacing human intuition with sophisticated quantitative tools.

This evolving technology of the future may be necessary to delve into the personal secrets of the past.

Source: popularmechanics.com/science/archaeology/

Similar Ads

Be the first to comment

your comment