# dailenson/SDT

This repository is the official implementation of Disentangling Writer and Character Styles for Handwriting Generation (CVPR 2023)

Repository: https://github.com/dailenson/SDT
Canonical: https://ross.abutalabs.com/products/sdt
Language: Python
License: MIT
License Family: permissive
Topics: deep-learning, handwriting-generation, transformer, generative-models, multimodal, pytorch-implementation, computer-vision, contrastive-learning, gmm
Last push: 2025-06-26T09:16:45+00:00

## Health v2 (maintenance only)
Score: 43/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 28, release rhythm 35, longevity 90
- inputs: {"age_days": 1261, "days_push": 433, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1403, forks 113 (observed 2026-08-28T04:04:37.878889+00:00)

## What it is
Official PyTorch implementation of the CVPR 2023 paper 'Disentangling Writer and Character Styles for Handwriting Generation' (SDT). It generates online handwriting with conditional content and style by disentangling writer-wise and character-wise style representations using a style-disentangled Transformer.

## Use cases
- generate handwriting in a specific person's style
- synthesize online Chinese handwriting with stroke order
- disentangle writer style from character style in handwriting
- generate offline Chinese handwriting images
- research handwriting generation models
- create custom-styled handwritten text samples

## When to choose
- you need to generate handwriting imitating a specific writer's style
- you want a research-grade model for online handwriting with stroke order
- you need Chinese handwriting synthesis with disentangled style control

## When to avoid
- you need production-ready handwriting OCR or recognition
- you need full-line English text generation without adaptation
- you want a plug-and-play tool with no ML setup

## Facets
- artifact type: library
- maturity: active
- function: deep-learning, machine-learning, image-processing, data-generation
- domain: computer-vision, deep-learning
- platform: python
- tags: handwriting-generation, style-disentanglement, transformer, contrastive-learning, cvpr-2023, chinese-handwriting, online-handwriting, research-code, natural-language-processing, linux, gpu

## Member repositories
- dailenson/SDT (main) score 43

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:37.878889+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:38:54.751950+00:00, confidence not recorded.
  - readme: https://github.com/dailenson/SDT (fetched 2026-08-28T04:04:37.878889+00:00, sha bfa3141cdf7e)
- Data as of 2026-08-30T08:39:29.467469+00:00.