Research map / 3D vision / NeRF and radiance fields
Depth-Supervised NeRF: Fewer Views and Faster Training for Free
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 COLMAP keypoint depth supervision with KL divergence | NeRF · NeRF Real, 2-view | PSNR: 13.5 -> 20.2 | Helped |
| Add COLMAP keypoint depth supervision with KL divergence | NeRF · DTU, 3-view | PSNR: 9.9 -> 16.9 | Mixed |
| Add COLMAP keypoint depth supervision with KL divergence | NeRF · DTU, 6-view | PSNR: 18.6 -> 20.6 | Helped |
| Add COLMAP keypoint depth supervision with KL divergence | NeRF · Redwood-3dscan, 2-view | PSNR: 10.5 -> 18.1 | Helped |
| Replace sparse COLMAP depth supervision with dense RGB-D depth (ablation) | DS-NeRF · Redwood-3dscan | PSNR: 18.1 -> 20.3; 22.9 -> 23.4; 23.8 -> 23.9 | Mixed |
| Use KL divergence rather than MSE for the depth-supervision loss (ablation) | DS-NeRF · NeRF Real, 2-view | PSNR: 19.5 -> 20.2 | Helped |
| Add depth supervision when fine-tuning IBRNet | IBRNet · NeRF Real, 2-view | PSNR: 14.4 -> 19.3 | Helped |
| Add depth supervision when fine-tuning MVSNeRF | MVSNeRF · NeRF Real, 5-view | PSNR: 21.8 -> 22.0 | No clear effect |
| Add depth supervision to NeRF | NeRF · NeRF Real, training speed to a given test PSNR | training iterations: 2-3x less training iterations | Helped |
| Use dense RGB-D depth supervision (ablation) | DS-NeRF · Redwood-3dscan, 2-view | Depth error%: 11.42 -> 5.81 | Helped |
Closest papers by meaning
- Rad-NeRF: Ray-decoupled Training of Neural Radiance Field (NeurIPS 2024 · similarity 0.92)
- AltNeRF: Learning Robust Neural Radiance Field via Alternating Depth-Pose Optimization (AAAI 2024 · similarity 0.91)
- Joint Optimization of Neural Radiance Fields and Continuous Camera Motion from a Monocular Video (CVPR 2025 · similarity 0.91)
- NeRFPrior: Learning Neural Radiance Field as a Prior for Indoor Scene Reconstruction (CVPR 2025 · similarity 0.91)
- CF-NeRF: Camera Parameter Free Neural Radiance Fields with Incremental Learning (AAAI 2024 · similarity 0.90)
- Mask-Based Modeling for Neural Radiance Fields (ICLR 2024 · similarity 0.90)
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.