Quoted by ISNA; Classical study scieists have created a new tool based on artificial ielligence that produces images of ancie times with more historical accuracy. Their approach provides not only more logical images, but also new opportunities to explore and understand the ancie world.
Felix Maier, a professor of ancie history, accompanied by Phillip Strewbel, a computational linguist from the University of Zurich (UZH). In collaboration with each other, the two have created a renovation platform called “Re-Experience History” that uses artificial ielligence to produce images with historical information from ancie Greece and Roman.
The platform iegrates three models of artificial ielligence that are trained based on academic literature and ancie sources to produce images that are rooted in scieific evidence. Artificial ielligence images often rely on modern and digital aesthetics, which can lead to ridiculous inconsistencies.
From ancie Rome to realistic illustration
The design of this platform is similar to the ChatGpt chat, but it also comes with other features. Users can choose from three image production models, each offering differe strengths. They can also automatically correct requests and set the input text, which is an image creation guide. This tool is iended for use in studies, education and museums.
Meyer says:
This platform allows historians to develop new perspectives. If we wa to create a picture of a victory eve in Rome, we have to ask ourselves accurate questions. How much was the conqueror visible among the crowded crowd? How was the victory displayed to make it effective and what path did the parade follow? Artificial ielligence may use today’s parade images. Pictured people may also be very muscular and shaved, and the ancie Roman city may look very clean.
“It was not easy to show the city and to have ordinary people,” said Stroller. The new platform, whose artificial ielligence is fed by accurate reconstruction of scieists’ work, can now create much more credible and more accurate images.
This approach was effective from the beginning. The team called on studes of ancie history to test differe models of artificial ielligence. According to studes’ feedback, the tool made them see Roman victories in a completely new view. The team hopes to achieve similar results through planned museums so that visitors can use the platform to produce their images based on selected exhibition topics. In line with collaborative practices, they can create small exhibitions inside large exhibitions.
According to them, even if the images produced are not accurate in all the historical details, these inaccuracies and defects can provide valuable educational experiences. The team’s approach creates new questions, even if the photos are not perfectly produced.
