Virtual University Journals
Search
Search
AI for Cultural Heritage: Advances in the Digital Reconstruction of Art and Artifacts
Muzaffar Ali Khan
Abstract:
Indus Valley Civilization, the world's oldest known urban civilization, left a cryptic script behind which is yet to be deciphered, owing mainly to the fragmented remains and lack of bilingual samples. Deciphering this old script has remained the problem of archaeologists and linguists for years. Recent progress in artificial intelligence, particularly computer vision and machine learning, has been instrumental in deciphering these complex symbols and inscriptions during recent years. This paper offers a technical overview of deep learning and image processing methods used in Indus seal and other epigraphic or grapheme-bearing objects studies. While models currently available show potential for promising identification and reconstruction of old texts, they are compromised by the scope and consistency of available data, and the quality and nuance of the inscriptions. Generative AI has the potential in the future to close these gaps and enable better restoration of removed or destroyed symbols and more nuanced understanding of the linguistic and cultural meaning of the Indus script.
Keywords:
Cultural Heritage, Artificial Intelligence, Digital Reconstruction , Artifacts
Full Text:
PDF