The DECRYPT project, a collaboration of linguists and computer scientists, aims to automate the process of cracking historical ciphers.
The CrypTool 2 software developed by Nils Kopal at Germany's University of Siegen has helped the DECRYPT researchers crack more than 2,500 ciphers.
Researchers can input cipher symbols into the software, which performs automated cryptanalysis to identify the homophones.
However, getting computer vision artificial intelligence (AI) to recognize and digitize enciphered texts, an important part of automating the decryption process, has proved more challenging.
A team led by Alicia Fornés at Spain's Autonomous University of Barcelona is testing AI models for this task, but variable handwriting in the old letters and the fact that the symbols often touch each other have complicated its work.
From New Scientist
View Full Article - May Require Paid Subscription
Abstracts Copyright © 2023 SmithBucklin, Washington, D.C., USA
No entries found