Would it be possible to share the training view camera poses
#6
by
RBJin
- opened
Thanks for the great dataset!
Would it be possible to share the camera poses (trajectories) for the training views?
I'm wondering if using these poses could potentially lead to higher-quality RGB and depth rendering — or if the current renderings are already optimized.
Sorry, we did not save the training data of the 3DGS (camera poses and rendering views)
RongWei-at
changed discussion status to
closed