Factorize detection merge #51
Conversation
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
cleherny
left a comment
cleherny
left a comment
There was a problem hiding this comment.
Thanks for your review. It was sometimes difficult for my end of the week-fried brain to keep track of the changes, as some parts of the code were moved to other scripts, but I hope I didn't miss anything important. In any case, it looks good!
I made a couple of minor comments. At the same time, I also answered to your comments of PR #49.
I haven't looked in details at the data_augmentation part, so perhaps @acerioni could have a deeper look, along with all the changes of this PR.
I tested the quarry example and it works fine. I ran one test with the empty tiles and it seems fine. Have you fixed the issue of assigning empty tiles to another dataset?
This comment has been minimized.
This comment has been minimized.
Analysis Details3 Issues
Coverage and Duplications
Project ID: swiss-territorial-data-lab_object-detector_AYZ4zWIzr7JdaaSXwexc |
No description provided.