AP

[Yongsz]

Hello, author, when reproducing the code, I got the following results with regard to the evaluation indicators in simple, medium, and difficult cases. Why does the detection result on pedestrians appear to have higher evaluation indicators in simple cases? , Hope to get your answer, thank you.

[Owen-Liuyuxuan]

This model seems trained on the "test_split" -> that is going to submit to the website?
If you want to reproduce the validation result, you should train on the "chen_split".

https://github.com/Owen-Liuyuxuan/visualDet3D/blob/master/config/Yolo3D_example#L84