Hi Ultralytics team ,
Thank you for your repo GitHub - ultralytics/xview-yolov3: xView 2018 Object Detection Challenge: YOLOv3 Training and Inference.
The Xview1 Public dataset has clear (balanced ) splits for Train , Validation and Test. Yet Validation labels have not been provided for the data training by the Xview Team.
My question, is what did you’ll use as your validation data set during the training process ? Did you further split the available annotated ‘train’ data into train and val subsets ? I was unable to figure this out through your readme on the repo and hence the explicit question.
PS: I have also written to the official XView team regarding this concern on their data set & they have responded .I can share their response here if it would help
Your response would be much appreciated , thank you !!