Researchers Develop Automatic Text Recognition for Ancient Cuneiform Tablets
Researchers at Martin Luther University Halle-Wittenberg (MLU) and Mainz University of Applied Sciences have developed a groundbreaking system that uses artificial intelligence (AI) to automatically decipher ancient cuneiform tablets. These tablets, which are over 5,000 years old, are among the oldest surviving written records of humankind. However, their complex writing system and poor condition have made them challenging to decipher. The new AI system, based on 3D models of the tablets, surpasses previous methods in accurately recognizing and translating the symbols. This breakthrough technology will provide unprecedented access to a vast range of knowledge preserved in cuneiform tablets.
The researchers used 3D models of nearly 2,000 cuneiform tablets, including 50 from MLU’s collection, to train the AI software. Despite the tablets’ weathered state, the system achieved remarkable accuracy in recognizing the characters. We were surprised to find that our system even works well with photographs, which are actually a poorer source material, says Ernst StΓΆtzner, the developer of the AI system.
The AI system functions similarly to optical character recognition (OCR) software, converting the images of the cuneiform writing into machine-readable text. This conversion enables easier reading and searching through the text once it is computerized. The team anticipates that the software will evolve to decipher weathered inscriptions in other three-dimensional contexts, such as cemetery artifacts.
While the current prototype of the AI system can reliably discern symbols from two languages, a total of twelve cuneiform languages are known to exist. The researchers envision expanding the software’s capabilities to include the decipherment of these languages in the future. This breakthrough not only provides access to previously inaccessible ancient knowledge but also opens up exciting possibilities for further research and exploration into humankind’s past.
The development of automatic text recognition for cuneiform tablets marks a significant advancement in historical research. It allows scholars and historians to decode and study these invaluable artifacts, shedding light on various aspects of ancient civilizations. With estimated one million cuneiform tablets still in existence worldwide, the AI system offers unparalleled opportunities to explore our shared history and gain deeper insights into the past.