HuggingFace Daily Papers(社區熱門論文)·· 6 天前AI 評分36
遮罩並非物件:以體積監督實現 3D Gaussian 分割
The Mask Is Not the Object: Volumetric Supervision for 3D Gaussian Segmentation
AI 導讀
VODA 提出一種即插即用的體積監督方法,直接監督與物件相關的 Gaussians,突破傳統基於渲染的監督方式,以改善 3D Gaussian Splatting 的物件分割。研究亦提出 Removal-3D,透過移除物件評估分割在體積上的完整性。
正文
Are we really doing "3D" segmentation?
In 3D Gaussian Splatting, accurate 2D masks don't necessarily mean complete 3D object segmentation.
We introduce VODA, a plug-and-play volumetric supervision approach that directly supervises object-associated Gaussians beyond conventional rendering-based supervision.
VODA consistently improves existing 3DGS segmentation methods, and we introduce Removal-3D to evaluate volumetric completeness through object removal.
來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co