You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Excuse me, have you ever train s2anet on UCAS_AOD dataset? I have problems aligning the accuracy in UCAS_AOD_Benchmark which you mentioned in another issue. And my model have a low ap on "car" class , while "airplane" is normal.
Do you have any idea on this problem? I 've been frustrated finding the bug for days. Hope to get your opinion. Thanks!
PS: One abnormal thing I' ve noticed in results(car.txt) is that the number of car bboxes is much higher (car 3684>airplane 2023) than the number of airplane bboxes. It's so weird.
The text was updated successfully, but these errors were encountered:
Excuse me, have you ever train s2anet on UCAS_AOD dataset? I have problems aligning the accuracy in UCAS_AOD_Benchmark which you mentioned in another issue. And my model have a low ap on "car" class , while "airplane" is normal.
Do you have any idea on this problem? I 've been frustrated finding the bug for days. Hope to get your opinion. Thanks!
PS: One abnormal thing I' ve noticed in results(car.txt) is that the number of car bboxes is much higher (car 3684>airplane 2023) than the number of airplane bboxes. It's so weird.
The text was updated successfully, but these errors were encountered: