What is the current accuracy of the uploaded project and trained model respectively?

[ZhangGongjie]

As above.

Thank you!

[yangxue0827]

The mAP of trained model is around 57.5，and this repo can get 60 or so. @ZhangGongjie

[ZhangGongjie]

May I ask how did you perform data augmentation to keep data balance?

Please kindly advise. (･ω´･ )

[yangxue0827]

Actually, I did not do any data balance. @ZhangGongjie

[ZhangGongjie]

That's amazing......

Actually, I am currently working on this dataset. I did not perform data balancing and get extremely low accuracy on some object instances like 'ground-track-field' and 'helicopter'.

Besides, could you please kindly advise how did you fuse the extremely large objects like ground-track-fields and roundabouts? Usually these kind of objects are split into parts when performing cropping to DOTA dataset.(･ω´･ )

[yangxue0827]

You can refer to my data process code. here. @ZhangGongjie

[yangxue0827]

The size of crop image is 800*800 with overlap 200 both in height and width. @ZhangGongjie

[ZhangGongjie]

From my experience, even an overlap of 200 cannot guarantee such large object fall into a single image without splitting.

[yangxue0827]

Yes, It's hard to guarantee coverage all objects.

[ZhangGongjie]

非常感谢！共同进步！ ：）