High-Fidelity Document Stain Removal via A Large-Scale Real-World Dataset and A Memory-Augmented Transformer

TL;DR

This paper introduces StainDoc, a large-scale dataset for document stain removal, and proposes StainRestorer, a memory-augmented Transformer that effectively removes stains while preserving document details.

Contribution

The paper presents the first high-resolution dataset for document stain removal and a novel Transformer-based model with memory augmentation for improved stain removal performance.

Findings

StainRestorer outperforms existing methods on the StainDoc dataset.

The dataset includes over 5,000 paired stained and clean document images.

Memory-augmented Transformer effectively captures hierarchical stain features.

Abstract

Document images are often degraded by various stains, significantly impacting their readability and hindering downstream applications such as document digitization and analysis. The absence of a comprehensive stained document dataset has limited the effectiveness of existing document enhancement methods in removing stains while preserving fine-grained details. To address this challenge, we construct StainDoc, the first large-scale, high-resolution ($2145\times2245$) dataset specifically designed for document stain removal. StainDoc comprises over 5,000 pairs of stained and clean document images across multiple scenes. This dataset encompasses a diverse range of stain types, severities, and document backgrounds, facilitating robust training and evaluation of document stain removal algorithms. Furthermore, we propose StainRestorer, a Transformer-based document stain removal approach. StainRestorer employs a memory-augmented Transformer architecture that captures hierarchical stain representations at part, instance, and semantic levels via the DocMemory module. The Stain Removal Transformer (SRTransformer) leverages these feature representations through a dual attention mechanism: an enhanced spatial attention with an expanded receptive field, and a channel attention captures channel-wise feature importance. This combination enables precise stain removal while preserving document content integrity. Extensive experiments demonstrate StainRestorer's superior performance over state-of-the-art methods on the StainDoc dataset and its variants StainDoc\_Mark and StainDoc\_Seal, establishing a new benchmark for document stain removal. Our work highlights the potential of memory-augmented Transformers for this task and contributes a valuable dataset to advance future research.