Thread

1. Lerc · 1790845274 · focus · HN ↗

   I'm rather surprised that it doesn't take the z-buffer as an input. I would have thought that would have provided useful information, it's one of the more useful forms of contolnet.

   1. avaer · 1790845516 · focus · HN ↗

      Even relatively small RGB -> depth models are pretty good. Which kind of implies depth is well encoded in the RGB, and adding depth would not really reduce entropy, while costing bandwidth.

Unofficial Hacker News client; not affiliated with Y Combinator.