Skip to content. | Skip to navigation

Personal tools
Password reminder

Navigation

Personal tools
Password reminder
You are here: Home / Type & typography / Deep Neural Network in Typeface

Deep Neural Network in Typeface

Keitaro Sakamoto, Tetsuo Sakomura ATypI 2016 • Warsaw, Poland Academy of Fine Arts in Warsaw ASP

It is said that it has been approx. 9,000 years since human being started to write a letter. It all started from carving on the shell of turtle, and then the transition from writing with brushes on papers to tapping on keyboards with displays. Has the evolution ended or are we still in the progress? We made a standpoint that AI (Artificial Intelligence) can be the next possible evolutionary tool. AI is closely related to image recognition in use such as Improvement in Safety of Automatic Vehicle and Heightening Factory Work Efficiency. The unique characteristic of AI is that it studies by itself repeatedly and does some case studies for self-progress. How can we plant this idea to Type Design? To sort out a typeface out of millions of Latin Typefaces? Or, automatically form thousands of glyphs from a single Japanese Typeface? It could plant and experiment in various fields. We would like to show you the experiment of how AI (Deep Neural Network) can decipher very unique Japanese character and how can AI adapt to its diversity.

Navigation