Results 96 comments of Jian Ding

You should use the same python interpret for complile and run.

What is the image source that you use? Have you conducted the multi-scale inference?

First, model trained with DOTA-v1.5 should be more accurate than the DOTA-v1.0. Second, I think there is a domain shift between your images and DOTA images, which may decrease the...

Have you set the correct test json and images?

You could visualize the results on test set chips to see if anything wrong.

It should be ok now.

Also looking for the training code. My Email is [email protected], thanks.