Publication detail

Capturing Fonts in the Wild

KOLÁŘ, M. HRADIŠ, M. ZEMČÍK, P.

Original Title

Capturing Fonts in the Wild

English Title

Capturing Fonts in the Wild

Type

conference paper

Language

en

Original Abstract

Editing text in photographs requires the ability to find the same font, which is impossible in many settings, such as historical or manually painted text. We present a method of extracting the font from a single photographed word, without relying on the retrieval of similar fonts. A deep net extracts style information and constructs the font for all characters, enabling novel applications in image editing, font creation, and the addition of language-specific characters with diacritics to existing fonts. A qualitative user study shows that this method improves convincing font capture by over 500% over prior work.

English abstract

Editing text in photographs requires the ability to find the same font, which is impossible in many settings, such as historical or manually painted text. We present a method of extracting the font from a single photographed word, without relying on the retrieval of similar fonts. A deep net extracts style information and constructs the font for all characters, enabling novel applications in image editing, font creation, and the addition of language-specific characters with diacritics to existing fonts. A qualitative user study shows that this method improves convincing font capture by over 500% over prior work.

Keywords

Computing methodologies, Image processing

Released

25.11.2020

Publisher

Eurographics Association

Location

NEUVEDEN

ISBN

978-3-03868-124-3

Book

Smart Tools and Apps for Graphics - Eurographics Italian Chapter Conference

Edition

NEUVEDEN

Issue number

NEUVEDEN

URL

Documents

BibTex


@inproceedings{BUT168499,
  author="Martin {Kolář} and Michal {Hradiš} and Pavel {Zemčík}",
  title="Capturing Fonts in the Wild",
  annote="Editing text in photographs requires the ability to find the same font, which is
impossible in many settings, such as historical or manually painted text. We
present a method of extracting the font from a single photographed word, without
relying on the retrieval of similar fonts. A deep net extracts style information
and constructs the font for all characters, enabling novel applications in image
editing, font creation, and the addition of language-specific characters with
diacritics to existing fonts. A qualitative user study shows that this method
improves convincing font capture by over 500% over prior work.",
  address="Eurographics Association",
  booktitle="Smart Tools and Apps for Graphics - Eurographics Italian Chapter Conference",
  chapter="168499",
  doi="10.2312/stag.20201238",
  edition="NEUVEDEN",
  howpublished="online",
  institution="Eurographics Association",
  year="2020",
  month="november",
  pages="0--0",
  publisher="Eurographics Association",
  type="conference paper"
}