How to do the inference for scene editing using the released codes?

[TyroneLi]

Hi thanks for the great job!

I am curious how I can do the inference for scene editing using the released codes, like Fig. 5 in the main paper?

Thanks a lo.

[Ziyi-Xiong1]

Hi. Thank you for your feedback.
You can directly edit the coordinates of the dynamic point cloud. More specifically, since the point cloud is per-pixel aligned, you can edit the per-pixel aligned point cloud.

[shuangshuangguo]

@TyroneLi @Ziyi-Xiong1 请问具体怎么做呢