AI RESEARCH
GA-GS: Generation-Assisted Gaussian Splatting for Static Scene Reconstruction
arXiv CS.AI
•
ArXi:2604.04331v1 Announce Type: cross Reconstructing static 3D scene from monocular video with dynamic objects is important for numerous applications such as virtual reality and autonomous driving. Current approaches typically rely on background for static scene reconstruction, limiting the ability to recover regions occluded by dynamic objects. In this paper, we propose GA-GS, a Generation-Assisted Gaussian Splatting method for Static Scene Reconstruction. The key innovation of our work lies in leveraging generation to assist in reconstructing occluded regions.