Closed HuizaiVictorYao closed 1 year ago
Hi,
Appreciate your time in reading our work.
We are randomly choosing box predictions but guided by the probability estimated by KDE. Objects with higher OBCs values are assigned higher probability to be chosen.
I will delve into OBC-based downsampling from two perspectives:
I hope this addresses your question.
Regards, Zhuoxiao (Ivan)
I got it. Thanks for your prompt reply!
Hi, your excellent work has brought me great inspiration. After looking through your paper, I wonder how you perform downsampling in stage 2 in detail. I mean that if we want to "uniformly downsample" a set of pseudo labels to obtain a subset with a length of \hat{B}/t, we can just randomly choose \hat{B}/t box predictions from the original box prediction set, then what does inverse KDE do in this process? Or “uniform downsampling" is just a proper noun of 3d detection?
Thanks for your reply in advance.