Deep Kern – Efficient digital glyph pair arrangement in fonts using deep learning based perceptual models

Development of a neural perceptual model to create, improve, accelerate and automate the process of spatial glyph arrangement in digital fonts for latin and non-latin scripts, with a confidence estimate for the resulting metrics.

https://www.aramis.admin.ch/Grunddaten/?ProjectID=55378