dataset

[zaoruis]

Hello, thank you for your work. I have a question about the data division. Do you divide label and unlabel by script? Does unlabeled2017 contain only pictures? Does annotations include only two json files for train2017 and val2017?