The Mask Is Not the Object: Volumetric Supervision for 3D Gaussian Segmentation
Abstract
A plug-and-play objective, VODA, groups per-pixel contributors by local depth to supervise all Gaussians in selected depth clusters, improving volumetric segmentation and enabling evaluation via Removal-3D and Removal Depth Error.
3D Gaussian Splatting enables object-level operations on Gaussian primitives, yet existing segmentation methods are largely trained and evaluated through rendered views. This mismatch can produce accurate rendered masks from incomplete 3D selections since alpha-blended feature supervision concentrates learning on dominant contributors while leaving occluded or low-contribution Gaussians weakly supervised. We propose VODA, a plug-and-play supervision objective that exploits local depth structure to group per-pixel contributors and directly supervises all Gaussians in selected depth clusters, bypassing contribution-weighted gradient delivery. To evaluate volumetric completeness, we introduce Removal-3D, a synthetic object-removal benchmark, and the Removal Depth Error (RDE), a metric comparing predicted and ground-truth depth changes. Across four rendering-based baselines, VODA consistently improves volumetric segmentation on Removal-3D while maintaining or improving performance on LERF and ScanNet. Further analysis shows that VODA recovers lower-contribution Gaussians, while simple gradient reweighting and spatial propagation fail to reproduce comparable gains.
Community
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.
This is an automated message from the Librarian Bot. I found the following papers similar to this paper.
The following papers were recommended by the Semantic Scholar API
- RefineSplat: Reconstruction-Guided Object-Centric Gaussian Splatting for Open-Vocabulary Querying (2026)
- Geometry-Guided Semantic Reconstruction for 3D Instance Segmentation (2026)
- MaRO-GS: Mask-Robust Object-Centric Gaussian Splatting from Inconsistent Multi-view Masks (2026)
- PePESeg3D: Perception Prior Enhances Multi-Scale Segmentation for 3D Gaussian Splatting (2026)
- Hybrid Gaussians for Robust Open-Vocabulary 3D Segmentation with Multi-View Object Association and Boundary Refinement (2026)
- VLSplat: Vision-Language Guided Object-Centric 3D Gaussian Splatting via Scene Graph (2026)
- RigidGS: Geometry-Aware Object Decoupling for Rigid-Body Interaction in 3D Gaussian Splatting (2026)
Please give a thumbs up to this comment if you found it helpful!
If you want recommendations for any Paper on Hugging Face checkout this Space
You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: @librarian-bot recommend
This is an automated message from the ResearchStudio team.
We created an interactive ResearchStudio Reel for this paper. It includes a visual poster, a video, and a blog, all available for download in editable formats.
Open the ResearchStudio Reel →
Download all files from Hugging Face
Please give this comment a thumbs up if you find the Reel helpful!
Want to explore or create Reels for more papers? Visit the ResearchStudio demo.
Get this paper in your agent:
hf papers read 2611.21383 Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash Models citing this paper 0
No model linking this paper
Datasets citing this paper 0
No dataset linking this paper
Spaces citing this paper 0
No Space linking this paper
