First of all, thank you very much for making your excellent work available! It's very beneficial to me!
There are 4,737 images for testing and 35,263 images for training mentioned in the paper, I can get the labels corresponding to each image from ”FLICKR-AES_image_labeled_by_each_worker.csv“ , but I don't know which ones are used for training or testing.
Could you please share the division of the dataset?
So I really hope to get your help. Thank you very much.
First of all, thank you very much for making your excellent work available! It's very beneficial to me! There are 4,737 images for testing and 35,263 images for training mentioned in the paper, I can get the labels corresponding to each image from ”FLICKR-AES_image_labeled_by_each_worker.csv“ , but I don't know which ones are used for training or testing. Could you please share the division of the dataset? So I really hope to get your help. Thank you very much.