asking for 3D slicing

[Crazywangq]

Thank you for sharing your code with the community, it's a fantastic resource.
I have a question regarding the 3D slicing based on 2D image cues. Could you clarify if this process involves reconstructing a 3D model from the 2D image before slicing, or does it simply entail slicing on the 2D image itself, or perhaps a 2.5D image? Additionally, are the results of the slicing characterized by a 3D contour? In other words, can the object be rotated to view it from different angles?
Thank you for your time and help!

[Jumpat]

Hi, thanks for your interest! The obtained assets are 3D Gaussians. Before segmentation beginning, we need to pre-train a 3D-GS model. After segmentation, the segmented target can be viewed from different viewpoints. You can try the saga_gui.py.