Srijika: OpenType-Layout-Reusing Font Restyling for Nine Indic Scripts
Srijika restyles glyph outlines from template fonts with latent diffusion to produce 66 OpenType TTFs across nine Brahmic scripts while preserving shaping consistency.
Srijika produces installable OpenType fonts for nine Brahmic scripts, including Devanagari, Tamil, and Bengali, by restyling glyph outlines from shaping-complete template fonts rather than generating from scratch. It preserves the template's cmap, GSUB closure, and GPOS data under a documented metric policy, yielding 66 TTFs that all pass the OpenType Sanitizer with HarfBuzz and CoreText reproducing template glyph-ID sequences. A reference-conditioned latent diffusion model redraws glyphs in a style selected via Lipika, a retrieval index over roughly 650 open-license font families, with content gating and shaped-cluster verification. The report also contributes a negative-results catalogue and cautions that measured style movement depends on an internal embedding trained on held-out families.
- Reuses template cmap, GSUB closure, and GPOS data so outputs are complete valid fonts by construction.
- Produces 66 TTFs passing OpenType Sanitizer; HarfBuzz and CoreText reproduce template glyph-ID sequences.
- Lipika retrieval index over ~650 open-license families selects styles; diffusion redraws glyphs with verification fallback.
Full article251 words · extracted from huggingface.co · click to collapse
We present Srijika, a system for producing installable OpenType fonts for nine Brahmic scripts: Devanagari, Tamil, Bengali, Telugu, Kannada, Malayalam, Gujarati, Gurmukhi, and Odia. Rather than generating fonts from scratch, Srijika restyles glyph outlines from shaping-complete template fonts. It preserves the template's cmap and GSUB closure and its GPOS data under a documented metric policy, making every output a complete font by construction. This addresses a central challenge of Indic font generation: hundreds to thousands of conjuncts, half forms, and matra variants must remain mutually consistent under OpenType shaping. Srijika produces 66 TTFs: 57 curated presets and nine open-vocabulary showcase fonts. All pass the OpenType Sanitizer, while HarfBuzz and CoreText reproduce the template glyph-ID sequences on conjunct-heavy probes. A full-closure audit covering 80,915 glyphs and 54,812 anchors quantifies metric changes. Natural-language style selection uses Lipika, a retrieval index over approximately 650 open-license font families. A reference-conditioned latent diffusion model redraws template glyphs in the selected style, followed by content gating, harmonization, and shaped-cluster verification with fallback to template outlines. We evaluate against no-learning baselines. On diffusion-training-family-held-out SSIM gates, template copying outperforms generation on 50 of 56 faces. Style movement is measurable only with an internal same-model embedding whose training corpus includes the held-out families, so these results require caution. A learned baseline, independent style metric, and human study are outside this report's scope. Our contributions are the layout-reusing formulation and pipeline, its nine-script audit and benchmark, and a negative-results catalogue covering failed conditioning, objective choices, and data-hull limits of reference-guided restyling.
Text extracted automatically; images, tables and formatting may be missing. Original: https://huggingface.co/papers/2609.05661