# Uni-DocDiff: A Unified Document Restoration Model Based on Diffusion

arXiv: https://arxiv.org/abs/2508.04055 · HTML: https://mlsanity.com/paper/2508.04055/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 32 of its 76 references that are in the map.
- First posted on arXiv 6 Aug 2025.

## Builds on (most cited first, 30 of 32)

- [Denoising Diffusion Implicit Models](https://mlsanity.com/paper/2010.02502/) (2020)
- [Label-Efficient Semantic Segmentation with Diffusion Models](https://mlsanity.com/paper/2112.03126/) (2021)
- [SEED: Semantics Enhanced Encoder-Decoder Framework for Scene Text Recognition](https://mlsanity.com/paper/2005.10977/) (2020)
- [ProRes: Exploring Degradation-aware Visual Prompt for Universal Image Restoration](https://mlsanity.com/paper/2306.13653/) (2023)
- [Diff-Plugin: Revitalizing Details for Diffusion-based Low-level Tasks](https://mlsanity.com/paper/2403.00644/) (2024)
- [TextCtrl: Diffusion-based Scene Text Editing with Prior Guidance Control](https://mlsanity.com/paper/2410.10133/) (2024)
- [DE-GAN: A Conditional Generative Adversarial Network for Document Enhancement](https://mlsanity.com/paper/2010.08764/) (2020)
- [DocTr: Document Image Transformer for Geometric Unwarping and Illumination Correction](https://mlsanity.com/paper/2110.12942/) (2021)
- [Document Rectification and Illumination Correction using a Patch-based CNN](https://mlsanity.com/paper/1909.09470/) (2019)
- [DocDiff: Document Enhancement via Residual Diffusion Models](https://mlsanity.com/paper/2305.03892/) (2023)
- [CDistNet: Perceiving Multi-Domain Character Distance for Robust Text Recognition](https://mlsanity.com/paper/2111.11011/) (2021)
- [Multiview Rectification of Folded Documents](https://mlsanity.com/paper/1606.00166/) (2016)
- [Unifying Image Processing as Visual Prompting Question Answering](https://mlsanity.com/paper/2310.10513/) (2023)
- [Geometric Representation Learning for Document Image Rectification](https://mlsanity.com/paper/2210.08161/) (2022)
- [DocRes: A Generalist Model Toward Unifying Document Image Restoration Tasks](https://mlsanity.com/paper/2405.04408/) (2024)
- [TPSNet: Reverse Thinking of Thin Plate Splines for Arbitrary Shape Scene Text Representation](https://mlsanity.com/paper/2110.12826/) (2021)
- [SPIRE: Semantic Prompt-Driven Image Restoration](https://mlsanity.com/paper/2312.11595/) (2023)
- [All-in-one Multi-degradation Image Restoration Network via Hierarchical Degradation Representation](https://mlsanity.com/paper/2308.03021/) (2023)
- [Document Dewarping with Control Points](https://mlsanity.com/paper/2203.10543/) (2022)
- [Visual Text Processing: A Comprehensive Review and Unified Evaluation](https://mlsanity.com/paper/2504.21682/) (2025)
- [Don't Forget Me: Accurate Background Recovery for Text Removal via Modeling Local-Global Context](https://mlsanity.com/paper/2207.10273/) (2022)
- [Water-Filling: An Efficient Algorithm for Digitized Document Shadow Removal](https://mlsanity.com/paper/1904.09763/) (2019)
- [UVDoc: Neural Grid-based Document Unwarping](https://mlsanity.com/paper/2302.02887/) (2023)
- [First Creating Backgrounds Then Rendering Texts: A New Paradigm for Visual Text Blending](https://mlsanity.com/paper/2410.10168/) (2024)
- [UDoc-GAN: Unpaired Document Illumination Correction with Background Light Prior](https://mlsanity.com/paper/2210.08216/) (2022)
- [Intrinsic Decomposition of Document Images In-the-Wild](https://mlsanity.com/paper/2011.14447/) (2020)
- [Arbitrary Reading Order Scene Text Spotter with Local Semantics Guidance](https://mlsanity.com/paper/2412.10159/) (2024)
- [GDB: Gated convolutions-based Document Binarization](https://mlsanity.com/paper/2302.02073/) (2023)
- [LDP: Generalizing to Multilingual Visual Information Extraction by Language Decoupled Pretraining](https://mlsanity.com/paper/2412.14596/) (2024)
- [DocBinFormer: A Two-Level Transformer Network for Effective Document Image Binarization](https://mlsanity.com/paper/2312.03568/) (2023)

## Built on by (most cited first, 5 of 5)

- [DocPure: Prompt-Free Unified Document Restoration via Degradation-Aware Structure-Guided Wavelet Modulation](https://mlsanity.com/paper/2608.09536/) (2026)
- [DocIntent: Answerability-Guided Agentic Restoration for Real-World Document Visual Question Answering](https://mlsanity.com/paper/2608.29037/) (2026)
- [SyntheticDoc: A Large Synthetic Dataset for Document Unwarping and Illumination Correction](https://mlsanity.com/paper/2609.15503/) (2026)
- [ENet-GP: Unified Document Image Restoration](https://mlsanity.com/paper/2609.33758/) (2026)
- [DDVR-Net: Degradation-Aware Decoupling and Visibility-Guided Restoration for Historical Document Recognition](https://mlsanity.com/paper/2610.16956/) (2026)

## Abstract (arXiv)

Removing various degradations from damaged documents greatly benefits digitization, downstream document analysis, and readability. Previous methods often treat each restoration task independently with dedicated models, leading to a cumbersome and highly complex document processing system. Although recent studies attempt to unify multiple tasks, they often suffer from limited scalability due to handcrafted prompts and heavy preprocessing, and fail to fully exploit inter-task synergy within a shared architecture. To address the aforementioned challenges, we propose Uni-DocDiff, a Unified and highly scalable Document restoration model based on Diffusion. Uni-DocDiff develops a learnable task prompt design, ensuring exceptional scalability across diverse tasks. To further enhance its multi-task capabilities and address potential task interference, we devise a novel textbfPrior Pool, a simple yet comprehensive mechanism that combines both local high-frequency features and global low-frequency features. Additionally, we design the textbfPrior Fusion Module (PFM), which enables the model to adaptively select the most relevant prior information for each specific task. Extensive experiments show that the versatile Uni-DocDiff achieves performance comparable or even superior performance compared with task-specific expert models, and simultaneously holds the task scalability for seamless adaptation to new tasks.