DRG-Font: Dynamic Reference-Guided Few-shot Font Generation via Contrastive Style-Content Disentanglement

Rejoy Chakraborty, Prasun Roy, Saumik Bhattacharya, Umapada Pal · 2026

Read DRG-Font: Dynamic Reference-Guided Few-shot Font Generation via Contrastive Style-Content Disentanglement by Rejoy Chakraborty, Prasun Roy, Saumik Bhattacharya, Umapada Pal (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-13797.txt.

Description

Chakraborty and three colleagues generate fonts from a few reference glyphs with DRG-Font. Capturing complex styles from few exemplars is hard, and prior methods lose local characteristics. Their contrastive strategy splits glyphs into style and content parts. A Reference Selection module dynamically picks the best style reference from a candidate pool. Style and content blocks learn style and shape priors. A Multi-Fusion Upsampling Block merges reference style with target content into the final glyph. The method beats state-of-the-art approaches on visual and test benchmarks.

Catalog record

CREATORRejoy Chakraborty, Prasun Roy, Saumik Bhattacharya, Umapada Pal
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE40941
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.13797

Browse the full library · About the Tower of Babel