search

LittlePey / SFD

Sparse Fuse Dense: Towards High Quality 3D Detection with Depth Completion (CVPR 2022, Oral)
Apache License 2.0
264 stars 35 forks source link

How to evaluate the model in different distence and different occlusion degrees with KITTI? #6

Closed Raiden-cn closed 2 years ago

Raiden-cn commented 2 years ago

Hi,@LittlePey. I really love your work, the code and the paper. But I got a question like the title said. How to evaluate the model in different distence and different occlusion degrees with KITTI like table 8 in your paper? I have read you paper and code carefully,But I didn't see those codes about distence and occlusion. Sorry to propose this silly question.I have a bad coding level. :( Could you give me some advices or something else? Thank you!

LittlePey commented 2 years ago

Hi, you can re-define the difficulty when generating infos.

Zyq1216 commented 6 months ago

Have you resolved it yet？May I ask how you did it？@Raiden-cn

Feidashen1 commented 5 months ago

Have you resolved it yet？May I ask how you did it？@Raiden-cn

Hello, have you resolved it yet？ How do we evaluate the model in different distance ? I have known how to evaluate different occlusion degrees. Maybe we can discuss about it.

Raiden-cn commented 5 months ago

I'm sorry, I don't remember the details clearly. I modified the range of GT and prediction box in mAP evaluation to achieve the effect I wanted. @Zyq1216 @Feidashen1

Zyq1216 commented 5 months ago

I haven't solved it yet.@Feidashen1

Feidashen1 commented 5 months ago

I'm sorry, I don't remember the details clearly. I modified the range of GT and prediction box in mAP evaluation to achieve the effect I wanted. @Zyq1216 @Feidashen1 Thank you for your reply. It has been very helpful to me！

Karkers commented 2 months ago

I'd like to know exactly what to do to get the same kind of assessment as in the paper, by different distances