facebookresearch / consistent_depth

We estimate dense, flicker-free, geometrically consistent depth from monocular video, for example hand-held cell phone video.
MIT License
1.61k stars 236 forks source link

MIDAS V.3 update #45

Open semel1 opened 3 years ago

semel1 commented 3 years ago

Any plan to update to MIDAS V.3 with "dpt_large-midas-2f21e586.pt" model?