Binarizing Documents by Leveraging both Space and Frequency

TL;DR

This paper introduces a novel document binarization method leveraging Fast Fourier Convolutions to effectively model both local and global information, outperforming traditional convolutional models and requiring fewer parameters than Vision Transformers.

Contribution

The work proposes a new binarization approach using Fast Fourier Convolutions to better capture global context with fewer parameters than ViT-based models.

Findings

Effective in handling various degradations

Outperforms standard convolutional models

Requires fewer parameters than Vision Transformers

Abstract

Document Image Binarization is a well-known problem in Document Analysis and Computer Vision, although it is far from being solved. One of the main challenges of this task is that documents generally exhibit degradations and acquisition artifacts that can greatly vary throughout the page. Nonetheless, even when dealing with a local patch of the document, taking into account the overall appearance of a wide portion of the page can ease the prediction by enriching it with semantic information on the ink and background conditions. In this respect, approaches able to model both local and global information have been proven suitable for this task. In particular, recent applications of Vision Transformer (ViT)-based models, able to model short and long-range dependencies via the attention mechanism, have demonstrated their superiority over standard Convolution-based models, which instead struggle to model global dependencies. In this work, we propose an alternative solution based on the recently introduced Fast Fourier Convolutions, which overcomes the limitation of standard convolutions in modeling global information while requiring fewer parameters than ViTs. We validate the effectiveness of our approach via extensive experimental analysis considering different types of degradations.