The Future of Graphic Novel Translation: Fully Automated Systems

The paper represents an approach to completely automate the process of translating graphic novels such as comics. With this approach, the aim to significantly lower the barrier caused by linguistic differences. This should help both comic creators and readers. Comics being a combination of both text and images are far more complicated to translate as compared to texts such as novels. Unlike translation of text which can be seen as an ever growing representative of what Natural Language Processing (NLP) can do, comic translation has seen few research publications that tackle the issue. Most approaches are theoretical in nature and even fewer cover the entire process, sticking to automating certain portions of the process instead. To solve this problem, we must first break down the process into a series of tasks that can be handled separately. This approach will be modular in nature where certain steps can be swapped out for more effective ones, as and when they appear. As this project involves a variety of fields such as NLP, Artificial Neural Networks (ANN), and Optical Character Recognition (OCR), each step may use various artificial intelligence or computer vision based techniques. This approach is inspired by the process of translating comics by human which includes various professionals such as translators, cleaners, typesetters, editors, etc. Though inspired by this, it will not strictly be limited to this, as there are various unique challenges and opportunities that arise when using a computer instead of a human.