A Benchmark for Chinese-English Scene Text Image Super-resolution

TL;DR

This paper introduces a new Chinese-English scene text image super-resolution benchmark dataset, Real-CE, with a focus on restoring complex Chinese characters, and proposes an edge-aware learning method to improve reconstruction quality.

Contribution

The paper presents the first real-world Chinese-English STISR benchmark dataset and an edge-aware learning approach tailored for complex Chinese characters.

Findings

Edge-aware loss improves reconstruction of Chinese characters

Benchmark dataset enables evaluation of STISR models on Chinese texts

Proposed method outperforms existing models on Real-CE

Abstract

Scene Text Image Super-resolution (STISR) aims to recover high-resolution (HR) scene text images with visually pleasant and readable text content from the given low-resolution (LR) input. Most existing works focus on recovering English texts, which have relatively simple character structures, while little work has been done on the more challenging Chinese texts with diverse and complex character structures. In this paper, we propose a real-world Chinese-English benchmark dataset, namely Real-CE, for the task of STISR with the emphasis on restoring structurally complex Chinese characters. The benchmark provides 1,935/783 real-world LR-HR text image pairs~(contains 33,789 text lines in total) for training/testing in 2$\times$ and 4$\times$ zooming modes, complemented by detailed annotations, including detection boxes and text transcripts. Moreover, we design an edge-aware learning method, which provides structural supervision in image and feature domains, to effectively reconstruct the dense structures of Chinese characters. We conduct experiments on the proposed Real-CE benchmark and evaluate the existing STISR models with and without our edge-aware loss. The benchmark, including data and source code, is available at https://github.com/mjq11302010044/Real-CE.