![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
AI helps decipher Herculaneum scrolls!
Dec. 8th, 2023 02:40 pmThis is REALLY something!
For those of you not up on your Roman history, Herculaneum was a city near Pompeii that was likewise destroyed when the volcano of Vesuvius erupted in 79 AD. A villa in Herculaneum was discovered to have several scrolls that were relatively intact stored there. The problem is that 'relatively intact' meant completely unreadable, and if you try to unroll it, it crumbles into bits. It's basically a lump of charcoal. But it is actually a scroll!
Enter advanced scanning techniques and the very clever application of AI.
By combining multiple scanning techniques and layering them, they can pick out the pigment from the paper! And by adding in AI and some clever programming, a student has identified letters in the scans! From the article: "Hundreds of badly charred ancient Roman scrolls found in a Roman villa have long been believed to be unreadable, but a 21-year-old computer science student at the University of Nebraska-Lincoln has successfully read the first text hidden within one of the rolled-up scrolls using a machine learning model. The achievement snagged Luke Farritor a $40,000 First Letters prize from the Vesuvius Challenge, a collaboration between private entrepreneurs and academics offering a series of rewards for milestones in deciphering the scrolls.
A second contestant, Youssef Nader, received a smaller $10,000 First Ink prize for essentially being the second person to decipher letters in a scroll. The main prize of $700,000 will be awarded to the first person to read four or more passages from one of the scrolls by December 31, and the founders are optimistic that this goal is achievable in light of these most recent breakthroughs."
The article goes on to describe techniques used to 'unroll', scan and analyze ancient documents including Egyptian papyri and a Dead Sea scroll that predates 600 CE. The difference here is that the Herculaneum scrolls used a carbon-based ink, meaning it won't fluoresce the way other inks would. A carbon ink inside a carbon briquette of a scroll exposed to the pyroclastic flow from a volcano = carbon on carbon! Talk about a low contrast situation!
"Farritor, a SpaceX summer intern, decided to train his own machine learning model on those crackle patterns, improving his model with each new pattern found. The model eventually revealed the word "πορφυρας" meaning "purple dye" or "cloths of purple," a word that rarely shows up in ancient texts. “When I saw the first image, I was shocked,” Federica Nicolardi, a papyrologist at the University of Naples in Italy who was among those who reviewed the findings, told Nature. “It was such a dream. I can actually see something from the inside of a scroll.”"
Machine Language Models, MLMs, and Large Language Models, LLMs, are broadly-speaking in the same field as AI. As I've said before, AI (the broad field) is going to be a very disruptive technology. But as can be seen here, it will clearly have some very beneficial applications and will help many fields such as archeology. These scientists will be able to revisit artifacts and look at them in ways that were unimaginable just a decade ago, much less a century ago.
https://arstechnica.com/science/2023/10/ai-helps-decipher-first-text-of-unreadable-ancient-herculaneum-scroll/
For those of you not up on your Roman history, Herculaneum was a city near Pompeii that was likewise destroyed when the volcano of Vesuvius erupted in 79 AD. A villa in Herculaneum was discovered to have several scrolls that were relatively intact stored there. The problem is that 'relatively intact' meant completely unreadable, and if you try to unroll it, it crumbles into bits. It's basically a lump of charcoal. But it is actually a scroll!
Enter advanced scanning techniques and the very clever application of AI.
By combining multiple scanning techniques and layering them, they can pick out the pigment from the paper! And by adding in AI and some clever programming, a student has identified letters in the scans! From the article: "Hundreds of badly charred ancient Roman scrolls found in a Roman villa have long been believed to be unreadable, but a 21-year-old computer science student at the University of Nebraska-Lincoln has successfully read the first text hidden within one of the rolled-up scrolls using a machine learning model. The achievement snagged Luke Farritor a $40,000 First Letters prize from the Vesuvius Challenge, a collaboration between private entrepreneurs and academics offering a series of rewards for milestones in deciphering the scrolls.
A second contestant, Youssef Nader, received a smaller $10,000 First Ink prize for essentially being the second person to decipher letters in a scroll. The main prize of $700,000 will be awarded to the first person to read four or more passages from one of the scrolls by December 31, and the founders are optimistic that this goal is achievable in light of these most recent breakthroughs."
The article goes on to describe techniques used to 'unroll', scan and analyze ancient documents including Egyptian papyri and a Dead Sea scroll that predates 600 CE. The difference here is that the Herculaneum scrolls used a carbon-based ink, meaning it won't fluoresce the way other inks would. A carbon ink inside a carbon briquette of a scroll exposed to the pyroclastic flow from a volcano = carbon on carbon! Talk about a low contrast situation!
"Farritor, a SpaceX summer intern, decided to train his own machine learning model on those crackle patterns, improving his model with each new pattern found. The model eventually revealed the word "πορφυρας" meaning "purple dye" or "cloths of purple," a word that rarely shows up in ancient texts. “When I saw the first image, I was shocked,” Federica Nicolardi, a papyrologist at the University of Naples in Italy who was among those who reviewed the findings, told Nature. “It was such a dream. I can actually see something from the inside of a scroll.”"
Machine Language Models, MLMs, and Large Language Models, LLMs, are broadly-speaking in the same field as AI. As I've said before, AI (the broad field) is going to be a very disruptive technology. But as can be seen here, it will clearly have some very beneficial applications and will help many fields such as archeology. These scientists will be able to revisit artifacts and look at them in ways that were unimaginable just a decade ago, much less a century ago.
https://arstechnica.com/science/2023/10/ai-helps-decipher-first-text-of-unreadable-ancient-herculaneum-scroll/
