Clarification of architecture used to do benchmarking on RGB dataset

[charan1561]

Hi @fangchangma @AkariAsai @abdo-eldesokey,

Thank you for sharing the benchmarking results on NYU Depth V2. I noticed the metrics for "Ours-RGB" and would like to understand more about the architecture or model configuration used to achieve these results.

Specifically, I am interested in the Backbone, Decoder and sample rate used for depth estimation model.

Thank you for your assistance!