AI analysis of dinosaur tracks suggests new interpretation • The Register

Paleontologists believe they’ve shed new light on a debate about what type of dinosaur might have left the ancient footprints at Australia’s Lark Quarry Conservation Park – by analyzing the footprints using AI.

The footprints on the ground have been interpreted by experts as fossilized footprints left when a group of about 150 dinosaurs evaded a predator about 93 million years ago – during the middle Cretaceous period. They have been preserved, protected and turned into a tourist attraction.

“Large dinosaur footprints were first discovered in the 1970s at a trail called Dinosaur Stampede National Monument, and for many years were believed to have been left by a predatory dinosaur such as Australovenatorwith legs nearly six feet long,” explained Anthony Romilio, a research fellow at the University of Queensland, Australia, who studied the tracks.

The patterns consist of a series of small indentations surrounded by a series of larger indentations that look like they were made by dinosaur footprints. However, paleontologists disagree on what type of creature might have created the larger track. The Stampede idea envisions a larger, carnivorous theropod dinosaur preying on the smaller ornithic herbivorous species.

But that may be wrong, according to an article published in the Journal of the Royal Society Interface.

Interpreting dinosaur tracks is difficult. The outlines of the prints are often blurred and evidence is subjective. A team of researchers from the University of Queensland and Liverpool John Moores University in the UK, along with colleagues in Germany, turned to machine learning for help.

“Some very different dinosaurs — carnivorous theropods and herbivorous ornithopods — had three-toed feet,” Romilio said The registry. “Distinguishing which dinosaur made the three-toed dinosaur tracks can be a source of frustration. Typical approaches use very general identifiers (long, narrow-toed theropods and short, chubby-toed ornithopods) and measurements of footprint landmarks (footprint is longer than wide for theropods, wider than long for ornithopods). But these still have trouble distinguishing tracks between these extreme forms.”

The researchers trained a convolutional neural network on 1,500 dinosaur footprints, which were divided into theropods, or ornithopods — a type of dinosaur closely related to ornithischians. The model only looked at the overall shape of the footprint images and did not consider other types of information – like their size or surface features. It was then tested on a series of 36 tracks.

The neural network spat out a score between 0 and 1 — a measure of how confident it was that an image should be classified as a theropod or ornithopod. The researchers acknowledged in their work that the small sample sizes of both the training and test datasets limit the performance of the system.

Still, they claimed that the model outperformed human experts during testing. “86 percent of the tracks were classified correctly, while 14 [percent] were misclassified. If all neural network decisions above 0.4 and below 0.6 are treated as ambiguous (22 [percent] of traces), classifies the neural network 67 [percent] of the traces correctly and 11 [percent] not correct. The human experts gave an average rating of 57 [percent] right, 20 [percent] wrong and 24 [percent] as ambiguous,” they wrote.

When they applied the model to the images of Dinosaur Stampede National Monument, the software classified the footprints as belonging to an ornithopod. “All but one of these tracks have been clearly classified as tracks from an ornithopod dinosaur – our prehistoric ‘predator,'” said Jens Lallensack, lead author at Liverpool John Moores University.

In other words, the larger dinosaur probably wasn’t a predator. The computer vision software is not perfect and has its own subjectivity due to its training. The authors warned, “It must be the job of the ichnologist” — a scientist who studies trace fossils rather than footprints than bones — “to combine the evaluation of shape by neural networks with all relevant contextual information to arrive at a meaningful interpretation of the track.” .”

Of course, the only way to definitively answer the question would be to restore the different dinosaur species from DNA trapped in amber and compare the tracks to contemporary specimens. Where’s John Hammond when you need him? ® AI analysis of dinosaur tracks suggests new interpretation • The Register

Rick Schindler

World Time Todays is an automatic aggregator of the all world’s media. In each content, the hyperlink to the primary source is specified. All trademarks belong to their rightful owners, all materials to their authors. If you are the owner of the content and do not want us to publish your materials, please contact us by email – The content will be deleted within 24 hours.

Related Articles

Back to top button