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
Short Summary - 3D Edge mapping can be cast as directly learning oriented points along the edges using supervision from multi-view 2D edge maps. Geometric regularization of 3D Gaussians for this task leads to a method that is efficiently trainable and produces state-of-the art results.
Sorry for the late response. We will have a small "hackaton" next Thursday. The goal is to collaborate to fix some issues and to get it updated. We will add your paper as well.
Short Summary - 3D Edge mapping can be cast as directly learning oriented points along the edges using supervision from multi-view 2D edge maps. Geometric regularization of 3D Gaussians for this task leads to a method that is efficiently trainable and produces state-of-the art results.
arXiv: https://arxiv.org/abs/2409.12886
Github: https://github.com/kunalchelani/EdgeGaussians
Might be a useful addition to the list?
The text was updated successfully, but these errors were encountered: