Ancient Textual Restoration Using Deep Neural Networks: A Literature Review

Ancient texts are important because they connect us with ancient civilizations, through which gain cultural, religious, and scientific knowledge, but these texts are often damaged over time due to environmental factors and the method of preservation that affects them, so the task of restoring these texts was one of the necessary tasks for specialists and researchers in ancient sciences. This paper provides an overview of the current state of research in this field. The review would examine the different approaches and methods used, the successes of these methods, and the challenges that need to be addressed in the future. The literature review would also cover the various types of datasets of ancient texts that have been restored using deep neural networks, such as manuscripts, inscriptions, and scrolls, as the different languages and the involved scripts.