# Improving Handwritten OCR with Training Samples Generated by Glyph Conditional Denoising Diffusion Probabilistic Model

arXiv: https://arxiv.org/abs/2305.19543 · HTML: https://mlsanity.com/paper/2305.19543/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Placed in Image generation, text-to-image and editing, one of 14 directions in Diffusion models.
- Builds on 36 of its 53 references that are in the map.
- First posted on arXiv 31 May 2023.

## Placed in

- Diffusion models / [Image generation, text-to-image and editing](https://mlsanity.com/diffusion/image-generation-text-to-image-and-editing/)

## Builds on (most cited first, 30 of 36)

- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [Denoising Diffusion Implicit Models](https://mlsanity.com/paper/2010.02502/) (2020)
- [Score-Based Generative Modeling through Stochastic Differential Equations](https://mlsanity.com/paper/2011.13456/) (2020)
- [Diffusion Models Beat GANs on Image Synthesis](https://mlsanity.com/paper/2105.05233/) (2021)
- [Deep Unsupervised Learning using Nonequilibrium Thermodynamics](https://mlsanity.com/paper/1503.03585/) (2015)
- [Hierarchical Text-Conditional Image Generation with CLIP Latents](https://mlsanity.com/paper/2204.06125/) (2022)
- [Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding](https://mlsanity.com/paper/2205.11487/) (2022)
- [A Style-Based Generator Architecture for Generative Adversarial Networks](https://mlsanity.com/paper/1812.04948/) (2018)
- [Classifier-Free Diffusion Guidance](https://mlsanity.com/paper/2207.12598/) (2022)
- [Generative Adversarial Networks](https://mlsanity.com/paper/2203.00667/) (2022)
- [Improved Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2102.09672/) (2021)
- [GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models](https://mlsanity.com/paper/2112.10741/) (2021)
- [FiLM: Visual Reasoning with a General Conditioning Layer](https://mlsanity.com/paper/1709.07871/) (2017)
- [Conditional Generative Adversarial Nets](https://mlsanity.com/paper/1411.1784/) (2014)
- [RePaint: Inpainting using Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2201.09865/) (2022)
- [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://mlsanity.com/paper/2109.10282/) (2021)
- [Pretraining is All You Need for Image-to-Image Translation](https://mlsanity.com/paper/2205.12952/) (2022)
- [Improved Noisy Student Training for Automatic Speech Recognition](https://mlsanity.com/paper/2005.09629/) (2020)
- [Self-Training for End-to-End Speech Recognition](https://mlsanity.com/paper/1909.09116/) (2019)
- [Pay Attention to What You Read: Non-recurrent Handwritten Text-Line Recognition](https://mlsanity.com/paper/2005.13044/) (2020)
- [ScrabbleGAN: Semi-Supervised Varying Length Handwritten Text Generation](https://mlsanity.com/paper/2003.10557/) (2020)
- [GANwriting: Content-Conditioned Generation of Styled Handwritten Word Images](https://mlsanity.com/paper/2003.02567/) (2020)
- [Evaluating Sequence-to-Sequence Models for Handwritten Text Recognition](https://mlsanity.com/paper/1903.07377/) (2019)
- [Rethinking Text Line Recognition Models](https://mlsanity.com/paper/2104.07787/) (2021)
- [Adversarial Generation of Handwritten Text Images Conditioned on Sequences](https://mlsanity.com/paper/1903.00277/) (2019)
- [Handwriting Transformers](https://mlsanity.com/paper/2104.03964/) (2021)
- [Learn to Augment: Joint Data Augmentation and Network Optimization for Text Recognition](https://mlsanity.com/paper/2003.06606/) (2020)
- [Text and Style Conditioned GAN for Generation of Offline Handwriting Lines](https://mlsanity.com/paper/2009.00678/) (2020)
- [Content and Style Aware Generation of Text-line Images for Handwriting Recognition](https://mlsanity.com/paper/2204.05539/) (2022)
- [Handwriting Recognition in Low-resource Scripts using Adversarial Learning](https://mlsanity.com/paper/1811.01396/) (2018)

## Built on by (most cited first, 8 of 8)

- [One-Shot Diffusion Mimicker for Handwritten Text Generation](https://mlsanity.com/paper/2409.04004/) (2024)
- [VATr++: Choose Your Words Wisely for Handwritten Text Generation](https://mlsanity.com/paper/2402.10798/) (2024)
- [Zero-Shot Paragraph-level Handwriting Imitation with Latent Diffusion Models](https://mlsanity.com/paper/2409.00786/) (2024)
- [Lyapunov Probes for Hallucination Detection in Large Foundation Models](https://mlsanity.com/paper/2603.06081/) (2026)
- [PLATTER: A Page-Level Handwritten Text Recognition System for Indic Scripts](https://mlsanity.com/paper/2502.06172/) (2025)
- [Advancing Offline Handwritten Text Recognition: A Systematic Review of Data Augmentation and Generation Techniques](https://mlsanity.com/paper/2507.06275/) (2025)
- [Quo Vadis Handwritten Text Generation for Handwritten Text Recognition?](https://mlsanity.com/paper/2508.09936/) (2025)
- [Dual Orthogonal Guidance for Robust Diffusion-based Handwritten Text Generation](https://mlsanity.com/paper/2508.17017/) (2025)

## Abstract (arXiv)

Constructing a highly accurate handwritten OCR system requires large amounts of representative training data, which is both time-consuming and expensive to collect. To mitigate the issue, we propose a denoising diffusion probabilistic model (DDPM) to generate training samples. This model conditions on a printed glyph image and creates mappings between printed characters and handwritten images, thus enabling the generation of photo-realistic handwritten samples with diverse styles and unseen text contents. However, the text contents in synthetic images are not always consistent with the glyph conditional images, leading to unreliable labels of synthetic samples. To address this issue, we further propose a progressive data filtering strategy to add those samples with a high confidence of correctness to the training set. Experimental results on IAM benchmark task show that OCR model trained with augmented DDPM-synthesized training samples can achieve about 45% relative word error rate reduction compared with the one trained on real data only.