what is the difference between those files ?

[mathshangw]

Thanks for your project but excuse me i have a question ..
1- what is the features expected needed from images
2- does object detection only is enough ? i mean i have the file you wrote with format : x1, y1, x2, y2, class_id, bbox_score only
can i make captioning with this file only ?
what is the difference between
features_flickr_train.pkl and flickr_detection_train.pkl