Introduction
The Aeneas team has introduced the first model designed for contextualizing ancient inscriptions, aimed at helping historians better interpret, attribute, and restore fragmentary texts. Writing was ubiquitous in the Roman world, from imperial monuments to everyday objects, encompassing political graffiti, love poems, and business transactions, providing rich insights into the diversity of daily life.
Background and Functionality of Aeneas
Often, these texts are fragmentary, weathered, or deliberately defaced. Restoring, dating, and placing them is nearly impossible without contextual information. The launch of Aeneas significantly accelerates this complex and time-consuming work, allowing historians to interpret texts by retrieving parallels from thousands of Latin inscriptions in seconds.
Aeneas is adaptable to other ancient languages, scripts, and media, helping to draw connections across a wider range of historical evidence. Co-developed with the University of Nottingham and in partnership with researchers from the Universities of Warwick, Oxford, and Athens University of Economics and Business, we aim to benefit as many people as possible by offering an interactive version of Aeneas for free.
Advanced Capabilities of Aeneas
Aeneas boasts several advanced functionalities:
- Parallels Search: It searches for parallels across a vast collection of Latin inscriptions, identifying historical connections to help historians contextualize inscriptions.
- Processing Multimodal Input: Aeneas is the first model to determine a text's geographical provenance using both text and images.
- Restoring Gaps of Unknown Length: For the first time, it can restore gaps in texts where the missing length is unknown, making it versatile for historians.
- State-of-the-Art Performance: Aeneas sets new benchmarks in restoring damaged texts and predicting when and where they were written.
How Aeneas Works
Aeneas is a multimodal generative neural network that takes both text and image of an inscription as input. Trained on the Latin Epigraphic Dataset (LED), consisting of over 176,000 inscriptions, it employs a transformer-based decoder to process the textual input, while specialized networks handle character restoration and geographical attribution.
Conclusion
Aeneas not only accelerates historians' work but also expands their perspectives, offering a new quantitative approach to engage with long-standing historical debates. Through collaboration with historians, Aeneas illustrates how AI can integrate with human expertise to advance the future of historical research.
Blogger's Review: The launch of Aeneas marks the dawn of a new era in historical research. The power of AI not only aids in text restoration and analysis but also provides historians with deeper insights. Its open-source and interactive version greatly enhances collaboration and knowledge sharing within the academic community, making it a noteworthy development!