mashey/dhivehi-image-text

Dhivehi图像-文本数据集是一个包含迪维希语（马尔代夫语）图像-文本对的数据集，用于机器学习和计算机视觉任务。数据集由10个批次（dv01-01至dv01-10）组成，每个批次包含训练、验证和测试分割，总图像数量为394,212张，平均每批约39,421张。分割比例为训练集80%、验证集10%、测试集10%。每个数据样本包括图像和文本两个特征，图像格式为图像类型，文本为字符串类型。数据集使用Apache 2.0许可证，任务类别为文本到图像，语言为迪维希语（dv），标签涉及OCR和视觉处理，规模类别为10万到100万之间。数据集可用于光学字符识别（OCR）和视觉语言模型训练等应用。Dhivehi Image-Text Dataset is a dataset of Dhivehi (Maldivian) image-text pairs for machine learning and computer vision tasks. It consists of 10 batches (dv01-01 to dv01-10), each with train, validation, and test splits, totaling 394,212 images with an average of approximately 39,421 images per batch. The split ratios are 80% for training, 10% for validation, and 10% for testing. Each data sample includes two features: image (in image format) and text (as string). The dataset is licensed under Apache 2.0, with task categories in text-to-image, language in Dhivehi (dv), tags including OCR and Vision, and size categories between 100K and 1M. It is suitable for applications such as optical character recognition (OCR) and vision-language model training.