Research map / 3D vision / 3D generation
FeatureNeRF: Learning Generalizable NeRFs by Distilling Foundation Models
Jianglong Ye, Naiyan Wang, Xiaolong 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 |
|---|---|---|---|
| add coordinate loss L_coord (ablation) | FeatureNeRF (DINO) · ShapeNet Chair 2D part co-seg | mIoU: 75.39 -> 76.55 | Improved |
| add coordinate loss L_coord (ablation) | FeatureNeRF (DINO) · ShapeNet Plane 2D part co-seg | mIoU: 72.83 -> 74.60 | Improved |
| use internal NeRF features instead of final output feature (ablation) | FeatureNeRF (DINO) · ShapeNet Chair 2D part co-seg | mIoU: 74.62 -> 76.55 | Improved |
| use internal NeRF features instead of final output feature (ablation) | FeatureNeRF (DINO) · ShapeNet Plane 2D part co-seg | mIoU: 72.29 -> 74.60 | Improved |
| distill DINO features into NeRF feature | NeRF Feature · ShapeNet Chair 2D part co-seg | mIoU: 72.02 -> 76.55 | Improved |
| distill DINO features into NeRF feature | NeRF Feature · ShapeNet Chair 3D part co-seg | mIoU: 65.23 -> 73.85 | Improved |
| distill diffusion features instead of DINO features | FeatureNeRF · ShapeNet Chair 2D part co-seg | mIoU: 76.55 -> 65.39 | Hurt |
| distill diffusion features into NeRF feature | NeRF Feature · ShapeNet Bottle 2D part co-seg | mIoU: 55.87 -> 53.91 | Hurt |
| distill DINO features (novel-view) | NeRF Feature · ShapeNet Bottle novel-view 2D part seg | mIoU: 87.14 -> 87.89 | No clear effect |
| distill diffusion features (novel-view) | NeRF Feature · ShapeNet Plane novel-view 2D part seg | mIoU: 69.45 -> 59.16 | Hurt |
| add feature distillation to PixelNeRF | PixelNeRF · ShapeNet Chair novel-view synthesis | PSNR: 23.29 -> 23.20 | No clear effect |
| FeatureNeRF (DINO) vs DINO baseline | DINO Baseline · ShapeNet Chair 3D part co-seg | mIoU: 52.67 -> 73.85 | Improved |
Closest papers by meaning
- TFS-NeRF: Template-Free NeRF for Semantic 3D Reconstruction of Dynamic Scene (NeurIPS 2024 · similarity 0.87)
- GP-NeRF: Generalized Perception NeRF for Context-Aware 3D Scene Understanding (CVPR 2024 · similarity 0.86)
- Diffusion 3D Features (Diff3F): Decorating Untextured Shapes with Distilled Semantic Features (CVPR 2024 · similarity 0.86)
- NeRF Analogies: Example-Based Visual Attribute Transfer for NeRFs (CVPR 2024 · similarity 0.86)
- DistillNeRF: Perceiving 3D Scenes from Single-Glance Images by Distilling Neural Fields and Foundation Model Features (NeurIPS 2024 · similarity 0.86)
- OpenNeRF: Open Set 3D Neural Scene Segmentation with Pixel-Wise Features and Rendered Novel Views (ICLR 2024 · similarity 0.86)
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.
