Can the network be used for cloud densification?

[Guru-Deep-Singh]

Dear Authors, I was planning on using your network for densification purposes. Thus, would like to use the resolution for the input and output as same i.e. channels are same. However, the input cloud is much less dense version of the output cloud. Therefore, the network would be forced to learn the mapping from sparse to dense representation. I would like to ask for your valuable feedback on the idea.

Thank you.