Details for Novel View Synthesis

[wrld]

Thanks for your great work! I want to ask about some implementation details of the novel view synthesis.
In your paper, you mentioned, "While the proposed guided optimization strategy needs the adapted depth priors as input to guide point sampling along the camera ray, we can still perform novel view synthesis by directly using the adapted depth priors from the nearest seen view.", but I can't find the corresponding code in your repo and wonder how to use depth priors from the nearest seen view.
Look forward to your reply.