The loss and degradation of cultural artifacts and ancient manuscripts due to conflict, natural disasters, and the passage of time pose significant challenges to global heritage preservation. This research explores the application of Generative Artificial Intelligence (Generative AI) in reconstructing and restoring incomplete or damaged cultural artifacts and ancient texts. Leveraging advanced deep learning models such as Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), and transformer-based language models, this approach enables the generation of high-fidelity restorations guided by historical records, image datasets, and linguistic corpora. The study outlines a multimodal framework that integrates visual and textual data to reconstruct eroded sculptures, faded paintings, and fragmented manuscripts. Preliminary results demonstrate that Generative AI can augment human expertise in digital heritage restoration, enhance public access to lost knowledge, and preserve cultural identity for future generations.
Wai Yie Leong (Mon,) studied this question.