Swahili Scene Text Detection and Recognition Dataset

本数据集名为Swahili Scene Text Detection and Recognition Dataset，由华中科技大学创建，旨在填补Swahili语言在场景文本检测与识别领域的研究空白。数据集包含976张来自坦桑尼亚城市及社交媒体的场景图像，涵盖商店标签、广告横幅等多种文本元素，每张图像均进行了手动词级标注。该数据集不仅为Swahili语言场景文本检测与识别提供了基准，也为未来研究提供了丰富的资源，特别适用于解决低资源语言在自然场景文本处理中的挑战。

This dataset, named Swahili Scene Text Detection and Recognition Dataset, was developed by Huazhong University of Science and Technology to fill the research gap in the field of scene text detection and recognition for the Swahili language. The dataset consists of 976 scene images collected from urban areas and social media platforms in Tanzania, covering various text elements such as store signs and advertising banners. Each image has been manually annotated at the word level. This dataset not only provides a benchmark for Swahili scene text detection and recognition, but also offers abundant resources for future research, and is particularly suitable for addressing the challenges faced by low-resource languages in natural scene text processing.