You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I wonder if you could explain the reason for improved tracking as per your paper. As I can see on a high level in your code the front-end (ORB-SLAM) and back-end (ORB-SLAM + 3DGS mapping) is structured so the 3DGS implementation does not modify / improve on the ORB map directly?
If this is the case the 3DGS and SLAM is essentially quite decoupled and what would be the reasoning of using the 3DGS mapping within the SLAM pipeline?
What am I missing?
Thanks
Best
Lucas
The text was updated successfully, but these errors were encountered:
Hi @liquorleaf @HuajianUP
First of all thanks for your impressive work!
I wonder if you could explain the reason for improved tracking as per your paper. As I can see on a high level in your code the front-end (ORB-SLAM) and back-end (ORB-SLAM + 3DGS mapping) is structured so the 3DGS implementation does not modify / improve on the ORB map directly?
If this is the case the 3DGS and SLAM is essentially quite decoupled and what would be the reasoning of using the 3DGS mapping within the SLAM pipeline?
What am I missing?
Thanks
Best
Lucas
The text was updated successfully, but these errors were encountered: