Task-driven real-world super-resolution of document scans

TL;DR

This paper presents a task-driven multi-task learning framework for super-resolution of document scans, improving OCR-related tasks by incorporating auxiliary high-level vision tasks and adaptive loss balancing.

Contribution

It introduces a novel multi-task training approach with auxiliary losses and dynamic weighting to enhance super-resolution for real-world document scans, focusing on OCR performance.

Findings

Improved text detection accuracy on real-world scans

Enhanced OCR performance with the proposed method

Better generalization from simulated to real-world data

Abstract

Single-image super-resolution refers to the reconstruction of a high-resolution image from a single low-resolution observation. Although recent deep learning-based methods have demonstrated notable success on simulated datasets -- with low-resolution images obtained by degrading and downsampling high-resolution ones -- they frequently fail to generalize to real-world settings, such as document scans, which are affected by complex degradations and semantic variability. In this study, we introduce a task-driven, multi-task learning framework for training a super-resolution network specifically optimized for optical character recognition tasks. We propose to incorporate auxiliary loss functions derived from high-level vision tasks, including text detection using the connectionist text proposal network, text recognition via a convolutional recurrent neural network, keypoints localization using Key.Net, and hue consistency. To balance these diverse objectives, we employ dynamic weight averaging mechanism, which adaptively adjusts the relative importance of each loss term based on its convergence behavior. We validate our approach upon the SRResNet architecture, which is a well-established technique for single-image super-resolution. Experimental evaluations on both simulated and real-world scanned document datasets demonstrate that the proposed approach improves text detection, measured with intersection over union, while preserving overall image fidelity. These findings underscore the value of multi-objective optimization in super-resolution models for bridging the gap between simulated training regimes and practical deployment in real-world scenarios.