BoundInk Treats Inter-Character Boundaries as Explicit Units, Cutting DTW by up to 47.8%
SUNGKYUNARCH · hf · 2026-09-28
BoundInk is a new framework for online handwriting generation. Existing methods rely on long-range sequence modeling and capture inter-character join, spacing, and alignment only implicitly, often producing plausible glyphs with broken cursive joins.
- Core idea: treat inter-character boundaries as explicit generation units, jointly modeling local transitions and surrounding text context to preserve writer style while improving line-level connectivity and spacing
- New eval: a boundary-aware evaluation framework assessing cursive continuity and spatial relationships beyond trajectory similarity
- Results: reduces normalized DTW by 17.6–47.8% across three benchmark-matched settings; preferred by humans in 78.0–82.6% of blind evaluation judgments
More from Research
- Cyber Index Alliance Launches With IBM, NVIDIA and Vercel to Standardize AI Cyber Defense Eval — ArtificialAnlys · 2026-09-28
- Artificial Analysis Details Cyber Index Methodology: Safety Refusals Score Zero, Tracked Separately — ArtificialAnlys · 2026-09-28
- Nissenbaum paper takes on privacy nihilism as AI inference erodes data-category frameworks — AllThingsApx · 2026-09-28
- Princeton researchers warn AI could slow science despite exploding paper output — AllThingsApx · 2026-09-28
- One-arm VR intervention during bimanual DAgger feels like an AI brain chip — neurosp1ke · 2026-09-28
- NeurIPS-rejected paper shows six agent dimensions to measure after benchmark saturation — random_walker · 2026-09-28