New light on the understanding of hieroglyphs thanks to the use of “Deep Learning”


New light on the understanding of the texts of ancient Egypt, even from simple photographic shots, thanks to the use of «Deep Learning», which uses neural network-based algorithms for image analysis. The application ofartificial intelligence it allows to classify hieroglyphs automatically – with very high accuracy and precision – regardless of the support on which they are written (papyrus, stone, wood).

This experimentation is the subject of a study published in the IEEE Access journal by Andrea Barucci e Costanza Cucci of the “Nello Carrara” Institute of Applied Physics of the National Research Council (Cnr-Ifac), Fabrizio Argenti e Marco Loschiavo of the Department of Information Engineering of the University of Florence, in collaboration with the Egyptologist Massimiliano Franci del Centro Studi CAMNES (Center for Ancient Mediterranean and Near Eastern Studies).

The application is the testimony of the happy union between methodologies used in the medical field and the human sciences. “Techniques based on deep neural networks now pervade all fields of knowledge.” explains Barucci of Cnr-Ifac and expert in biomedical image analysis with machine and deep learning techniques. “We asked ourselves if this paradigm could be translated into an apparently distant and different sphere, such as the recognition of ancient symbols. Our experience in the field of clinical images suggested that convolutional neural networks are extremely powerful and versatile tools, however the challenge was open ».

The research not only demonstrates the possibility of automatic translation of ancient Egyptian documents, but offers new perspectives for solving open questions such as the coding, recognition and transliteration of hieroglyphic signs. The use of artificial intelligence helps scholars to deepen different aspects of writing. «The topo-syntax of hieroglyphic signs combined to form words; linguistic analysis of texts; the recognition of corrupted, rewritten, canceled signs; up to the possibility of the recognition of the school of the scribe or the hand of the sculptor », continues the Egyptologist Massimiliano Franci.

«The intuition of the expert is still fundamental in the integration of the complex analyzes provided by artificial intelligence (AI) algorithms and the future requires an ever greater harmonization between computer and human analysis. Our study wants to highlight how the analysis tools based on AI can support investigations in the Egyptological field, integrating with the work of the archaeologist (human in the loop) ».

“This study was born from Marco Loschiavo’s thesis” – continues Fabrizio Argenti of the Department of Information Engineering of the University of Florence – “From the engineering point of view we were sure of the potential of the analysis tools chosen, however this was a bench important test, as the type of application is completely different. We wanted to explore a new area of ​​research, which proved to be extremely interesting and promising ».

Cnr-Ifac has a highly multidisciplinary character in its roots and skills. “In facilitating the exchange and cross-fertilization between different research fields, as happened for this work, skills in Egyptology, computer engineering and applied physics were combined”, adds Costanza Cucci, expert in data analysis in the Cultural Heritage field.

“The hope”, continues Barucci, “is that this first study will pave the way towards a stable collaboration between the communities that deal with archeology and artificial intelligence, to create new tools that facilitate the work of scholars of the scriptures of ancient civilizations” .


Source: RSS DiariodelWeb.it Innovazione by www.diariodelweb.it.

*The article has been translated based on the content of RSS DiariodelWeb.it Innovazione by www.diariodelweb.it. If there is any problem regarding the content, copyright, please leave a report below the article. We will try to process as quickly as possible to protect the rights of the author. Thank you very much!

*We just want readers to access information more quickly and easily with other multilingual content, instead of information only available in a certain language.

*We always respect the copyright of the content of the author and always include the original link of the source article.If the author disagrees, just leave the report below the article, the article will be edited or deleted at the request of the author. Thanks very much! Best regards!