From Generation to Matching: A Development Report on Personalized Chinese Handwriting

2026-08-16

AI Summary

This study addresses the conflict between structural regularity and stylistic fidelity in personalized Chinese handwriting generation by proposing a paradigm shift from synthetic generation to real-sample retrieval matching. We introduce a virtual writer composition framework grounded in human equivalence classes, integrating ink feature extraction, character-level percentile modeling, and a greedy hardest-first selection strategy to ensure precise assembly of style-compatible characters. Experimental results demonstrate that this approach achieves full matching for 197 target characters and 100% coverage on a hundred-character evaluation set. The generated outputs consistently attain Grade B quality, with select samples approaching Grade A authentic standards. Consequently, this method effectively resolves the inherent style-structure contradiction prevalent in few-shot scenarios, offering a robust solution for high-fidelity personalized handwriting synthesis.

Abstract

This paper documents a frozen engineering project on personalized Chinese handwriting. The project started from approximately 200 real handwriting images from one user, covering 197 unique Chinese characters, and was initially formulated as few-shot generation of unseen characters. A sequence of canonical-centered personalization routes repeatedly exposed the same conflict: increasing structural pressure made outputs more canonical, while increasing personalization could damage identity-defining strokes. The project was therefore reset around real-human character equivalence classes. A multi-writer CASIA candidate pool showed that a USER-compatible realization often already existed among valid human samples. The task consequently changed from synthesis to character-wise matching, followed by cross-writer composition into a virtual writer. The frozen system uses real-ink features, character-specific human population percentiles, top-20 candidate pruning, and greedy hardest-first whole-row selection. On the covered target set, all 197 USER characters had real-human candidates, and the 100-character evaluation subset was covered 100/100. Knowncharacter held-out comparisons included a row judged visually almost indistinguishable from genuine USER handwriting. A 60- episode stability audit placed every episode in a predefined A-like machine-proxy region, but these were not independent human A-level judgments. The final evidence supports stable practical B-level quality, with many outputs approaching A-level under the USER-defined criterion. The report records why generation became unnecessary for this case without claiming unrestricted or universal handwriting synthesis.

Problem

• Personalized Chinese Handwriting
• Few-shot Generation
• Style-Structure Conflict
• Character Matching

Innovation

• Character-wise Matching
• Cross-writer Composition
• Real-human Equivalence Classes
• Greedy Hardest-first Selection
• Personalized Chinese Handwriting