Decoding Ancient Scrolls of Herculaneum: Students Harness AI Technology
The Greek word meaning "purple" was extracted from the scrolls of Herculaneum vesuvius challenge Nearly 2,000 years after they were buried by the eruption of Mo Nearly 2,000 years after they were buried by Mount Vesuvius in 79 AD, scrolls in the library of the ancient Roman town of Herculaneum began to reveal their secrets. By studying 3D X-ray scans of the scrolls, researchers deciphered the word “porphyra” on one of them. Luke Faritor, a 21-year-old computer science student at the University of Nebraska-Lincoln, trained his AI to identify nearly invisible ink-like patterns in 3D scans. Yusef Nader, a data science graduate student at Germany’s Freie Universität Berlin, independently discovered the same word using another AI technique for detecting possible letter shapes in scrolling image segments. This discovery builds on the work of previous Vesuius Challenge contributors who designed computational tools to map scroll segments, and 3D imaging and computational techniques to digitally reveal the hidden contents of these scrolls. This advance could pave the way for someone to claim the Vesuvian Challenge grand prize, Read four passages from two intact scrolls by December 31, 2023 to win a prize worth $700,000.
ที่ตีพิมพ์ : 2 ปีที่แล้ว โดย admin ใน Tech Science
Nearly 2,000 years after they were buried by the eruption of Mount Vesuvius in 79 AD, scrolls in the library of the ancient Roman town of Herculaneum began to reveal their secrets. Tightly wrapped papyrus scrolls were charred in the disaster, and the nearby town of Pompeii was also destroyed. However, by studying 3D X-ray scans of the scrolls, researchers deciphered the word “porphyra”, meaning “purple”, on one of them.
The breakthrough was Luke Faritor, a 21-year-old computer science student at the University of Nebraska-Lincoln. His successes included training his AI to identify nearly invisible ink-like patterns in 3D scans.
“I was shocked when I saw Luke’s first words,” he says. michael mc oscar from the University of Cologne in Germany was not involved in the discovery. “It was immediately convincing. I was like, ‘Sir, this is Greek.'”
Farritor analyzed 3D scanning as an open source competitor. vesuvius challenge, which awards a series of awards for deciphering the scrolls. He submitted his findings in August.
Almost at the same time Yusef NaderThe data science graduate student at Germany’s Freie Universität Berlin independently discovered the same word using another AI technique for detecting possible letter shapes in scrolling image segments.This gave us an even clearer picture It’s a scrolling segment and already produces a new clear image of something else. McCosker called the snapshot of Nader’s first words “even more impressive” than Faritor’s.
This discovery builds on the work of previous Vesuvius Challenge contributors who designed computational tools to map scroll segments. This was also made possible thanks to 3D X-ray scans created by teams such as: brent shields at the University of Kentucky.
Papyrologists were once able to study the Herculaneum scrolls only by physically unrolling them, but in the process they inevitably damaged the ancient papyri, which had been charred by the heat of the volcanic debris that buried them. Oscar says. And even after researchers began using 3D imaging and computational techniques to digitally reveal the hidden contents of the scrolls, “attempts to read the papyrus while it was still rolled were a mirage.” “It was,” he says.
This latest advance could pave the way for someone to claim the Vesuvius Challenge grand prize, Read four passages from two intact scrolls by December 31, 2023 to win a prize worth $700,000.
“I’m confident that Luke, Youssef and the other contestants will be able to solve the entire role,” says McCosker. “So far, all the unrolled papyrus documents we are studying are missing beginnings and are in poor condition, so the prospect of being able to read the complete text from beginning to end is highly unlikely. It’s a big one.”
หัวข้อ: AI