Local Style Awareness of Font Images

TL;DR

This paper introduces an attention mechanism to identify important local parts of fonts, enhancing local style-aware font generation by focusing on style-relevant regions without manual annotation.

Contribution

It proposes a novel attention mechanism trained in a quasi-self-supervised manner to improve local style recognition and font generation quality.

Findings

Attention mechanism effectively identifies style-relevant local parts.

The new loss function improves font generation quality.

Enhances few-shot font generation models.

Abstract

When we compare fonts, we often pay attention to styles of local parts, such as serifs and curvatures. This paper proposes an attention mechanism to find important local parts. The local parts with larger attention are then considered important. The proposed mechanism can be trained in a quasi-self-supervised manner that requires no manual annotation other than knowing that a set of character images is from the same font, such as Helvetica. After confirming that the trained attention mechanism can find style-relevant local parts, we utilize the resulting attention for local style-aware font generation. Specifically, we design a new reconstruction loss function to put more weight on the local parts with larger attention for generating character images with more accurate style realization. This loss function has the merit of applicability to various font generation models. Our experimental results show that the proposed loss function improves the quality of generated character images by several few-shot font generation models.