-Person Search With Natural Language Description PROJECT-
DISCRIPTION
The data was collected at TU Rangsit. The data consists of over 24000 image-description pairs. The zip file contains a JSON file and another zip file. Each object in the JSON file contains a field that has the name of the corresponding image. The images can be found in the nested zip file.
As can be seen from the above image, the hilighted area is the data for the first image. The field img_id refers to the image file name, which is 000001.jpg in the image folder. The field captions is list of strings, where each string is a description for this image.