BoundInk: Boundary-Aware Online Handwriting Generation

Hugging Face Daily Papers Papers
handwriting-generation boundary-aware writer-conditioned sequence-modeling generative-model neural-networks

Summary

BoundInk introduces a writer-conditioned framework for online handwriting generation that explicitly models inter-character boundaries to improve cursive continuity and spatial alignment, outperforming existing methods with significant reductions in error and high human preference.

Realistic online handwriting depends not only on individual character shapes, but also on how a writer connects, spaces, and aligns adjacent characters. Existing methods rely primarily on long-range sequence modeling and capture these inter-character behaviors only implicitly. This often produces plausible glyphs accompanied by broken cursive joins, inconsistent spacing, or writer-inconsistent transitions. We introduce BoundInk, a writer-conditioned framework that treats inter-character boundaries as explicit generation units. By jointly modeling local transitions and surrounding text context, BoundInk preserves writer-specific glyph appearance while improving connectivity and spacing across complete text lines. We further introduce a boundary-aware evaluation framework that directly assesses cursive continuity and spatial relationships between characters beyond conventional trajectory similarity. Across three benchmark-matched settings, BoundInk improves all applicable boundary-quality measures and reduces normalized dynamic time warping by 17.6--47.8%. In blind human evaluations, BoundInk outputs are preferred in 78.0--82.6% of valid criterion-wise judgments.
Original Article
View Cached Full Text

Cached at: 09/28/26, 12:04 PM

Paper page - BoundInk: Boundary-Aware Online Handwriting Generation

Source: https://huggingface.co/papers/2604.02103

Abstract

Realisticonlinehandwritingdependsnotonlyonindividualcharactershapes,butalsoonhowawriterconnects,spaces,andalignsadjacentcharacters.Existingmethodsrelyprimarilyonlong-rangesequencemodelingandcapturetheseinter-characterbehaviorsonlyimplicitly.Thisoftenproducesplausibleglyphsaccompaniedbybrokencursivejoins,inconsistentspacing,orwriter-inconsistenttransitions.WeintroduceBoundInk,awriter-conditionedframeworkthattreatsinter-characterboundariesasexplicitgenerationunits.Byjointlymodelinglocaltransitionsandsurroundingtextcontext,BoundInkpreserveswriter-specificglyphappearancewhileimprovingconnectivityandspacingacrosscompletetextlines.Wefurtherintroduceaboundary-awareevaluationframeworkthatdirectlyassessescursivecontinuityandspatialrelationshipsbetweencharactersbeyondconventionaltrajectorysimilarity.Acrossthreebenchmark-matchedsettings,BoundInkimprovesallapplicableboundary-qualitymeasuresandreducesnormalizeddynamictimewarpingby17.6--47.8%.Inblindhumanevaluations,BoundInkoutputsarepreferredin78.0--82.6%ofvalidcriterion-wisejudgments.

View arXiv pageView PDFGitHub6Add to collection

Get this paper in your agent:

hf papers read 2604\.02103

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2604.02103 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2604.02103 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2604.02103 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

Inkling

Product Hunt

Inkling is an open-weights 975B multimodal AI model designed for fine-tuning.

Welcome Inkling by Thinking Machines

Hugging Face Blog

Inkling by Thinking Machines is a large open multimodal LLM with ~1T parameters, 1M context, and native support for image, audio, and text. It uses a Mixture-of-Experts architecture and is available on Hugging Face with day-0 inference support.

thinkingmachines/Inkling-NVFP4

Hugging Face Models Trending

Inkling is a 975B-parameter sparse mixture-of-experts multimodal model accepting text, image and audio inputs and generating text outputs. Released with open weights for research, fine-tuning, and integration.

thinkingmachines/Inkling

Hugging Face Models Trending

Inkling is a large open-weights multimodal model (975B total, 41B active parameters) using a sparse MoE architecture, accepting text, image, and audio inputs and generating text outputs, intended for agentic systems, coding assistants, and chatbots.