Research map / 3D vision / Gaussian splatting
Mastering Regional 3DGS: Locating, Initializing, and Editing with Diverse 2D Priors
Lanqing Guo, Yufei Wang, Hezhen Hu, Yan Zheng, Yeying Jin, Siyu Huang, Zhangyang Wang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use 3DGS with the proposed method instead of NeRF-based IN2N | editing speed · Text-guided 3D editing (three 3D scenes) | editing time: ~51min -> ~2min | Improved |
| compare against 3DGS-based editors | editing speed · Text-guided 3D editing (three 3D scenes) | editing time: up to 4x speedup compared to existing 3DGS-based methods | Improved |
| compare against NeRF-based editors | editing speed · Text-guided 3D editing (three 3D scenes) | editing time: 25x speedup compared to NeRF-based methods | Improved |
| proposed method vs DGE on fidelity | Ours · Text-guided 3D editing (three 3D scenes) | FID, IIS: FID 68.60 -> 65.96; IIS 0.8183 -> 0.8308 | Improved |
| proposed method vs IN2N on text alignment | Ours · Text-guided 3D editing (three 3D scenes) | CTIDS: CTIDS 0.2128 -> 0.2053 | Hurt |
| add 3D localization (ablation) | Ours · Text-guided 3D editing (three 3D scenes) | reconstruction quality (qualitative) | Improved |
| replace SAM-prompt localization with diffusion-noise-difference 3D localization (ablation) | Ours · Text-guided 3D editing (three 3D scenes) | localization (qualitative) | Improved |
| add depth-based initialization (ablation) | Ours · Text-guided 3D editing (three 3D scenes) | finetuning iterations: 1,000 iterations -> 300 iterations | Improved |
| add depth-based initialization (ablation) | Ours · Text-guided 3D editing (three 3D scenes) | multi-view consistency (qualitative) | Improved |
| use global editing with all-one mask instead of localized editing | Ours · Text-guided 3D editing (three 3D scenes) | finetuning iterations: 1,000-1,500 iterations -> 2,000 iterations over four cycles | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- Gaussctrl — GaussCtrl: Multi-View Consistent Text-Driven 3D Gaussian Splatting Editing (ECCV 2024) · better on 5 of 5 shared results
- ViCA-NeRF — ViCA-NeRF: View-Consistency-Aware 3D Editing of Neural Radiance Fields (NeurIPS 2023) · better on 3 of 4 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- EditSplat: Multi-View Fusion and Attention-Guided Optimization for View-Consistent 3D Scene Editing with 3D Gaussian Splatting (CVPR 2025 · similarity 0.91)
- 3DGS-Enhancer: Enhancing Unbounded 3D Gaussian Splatting with View-consistent 2D Diffusion Priors (NeurIPS 2024 · similarity 0.90)
- 3DGS-Drag: Dragging Gaussians for Intuitive Point-Based 3D Editing (ICLR 2025 · similarity 0.89)
- Feature 3DGS: Supercharging 3D Gaussian Splatting to Enable Distilled Feature Fields (CVPR 2024 · similarity 0.89)
- SceneSplat++: A Large Dataset and Comprehensive Benchmark for Language Gaussian Splatting (NeurIPS 2025 · similarity 0.89)
- GaussianPro: 3D Gaussian Splatting with Progressive Propagation (ICML 2024 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
