Bangla Image Captioning

We have developed this dataset for the Bangla image caption. Here, we recorded 1333 unique images with five captions each. All the files are in jpg format. This dataset contains a variety of images, including images obtained from DSLRs and phone cameras. The comprehensive nature of the dataset and diverse image sources make it a valuable resource for advancing Bengali picture captioning research. All images and captions ( xlsx ) are in the zip file and some image samples are in the images folder.本数据集专为孟加拉语图像字幕任务研发。本次共收录1333张独特图像，每张图像配有5条字幕标注。所有图像文件均为JPG格式。本数据集涵盖多元图像来源，包括数码单反相机（DSLR）与手机相机拍摄的图像。该数据集兼具全面性与多样化的图像来源特性，可为孟加拉语图像字幕研究的推进提供宝贵的资源支持。所有图像与字幕标注文件（xlsx格式）均打包于该压缩文件中，部分图像示例存放在images文件夹内。