Research map / 3D vision / NeRF and radiance fields
HyP-NeRF: Learning Improved NeRF Priors using a HyperNetwork
Read the paperCheck 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 |
|---|---|---|---|
| Generate both instance-specific MRHEs and NeRF MLP weights (ablation) | HyP-NeRF · ABO-512 Chairs and Sofas, single-view generation | PSNR: Chairs: 23.05 -> 24.23; Sofa: 23.54 -> 23.96 | Helped |
| Add denoising and finetuning to the predicted NeRF (ablation) | HyP-NeRF · ABO-512 Chairs | FID: 102.45 -> 68.11 | Helped |
| Add denoising and finetuning to the predicted NeRF (ablation) | HyP-NeRF · ABO-512 Sofa | LPIPS: 0.174 -> 0.18 | No clear effect |
| Add MRHE prediction instead of relying on positional encodings (ablation) | HyP-NeRF · ABO Chairs | PSNR: 26.42 -> 29.23 | Helped |
| Add MRHE prediction instead of relying on positional encodings (ablation) | HyP-NeRF · ABO Chairs | CD: 0.0100 -> 0.0075 | Helped |
| Use HyP-NeRF rather than CodeNeRF | Single-view NeRF generation · ABO-512 Chairs and Sofas | PSNR: Chairs: 19.86 -> 24.23; Sofa: 19.56 -> 23.96 | Helped |
| Compress instance NeRFs into a shared learned prior | HyP-NeRF versus InstantNGP · ABO Chairs compression | PSNR: 35.43 -> 31.37 | Hurt |
| Use a CLIP-based query network to retrieve training instances | HyP-NeRF · ABO Chairs and Sofa retrieval | Top 1 accuracy: 98.72%; 91.6% | Helped |
| Increase the number of input views for test-time optimization | HyP-NeRF · SRN Chairs | render quality | Mixed |
| Add denoising and finetuning to the predicted NeRF (ablation) | HyP-NeRF · ABO Chairs geometric consistency | CD: 0.0064 -> 0.0062 | No clear effect |
Closest papers by meaning
- HyperFields: Towards Zero-Shot Generation of NeRFs from Text (ICML 2024 · similarity 0.91)
- InsertNeRF: Instilling Generalizability into NeRF with HyperNet Modules (ICLR 2024 · similarity 0.91)
- LuSh-NeRF: Lighting up and Sharpening NeRFs for Low-light Scenes (NeurIPS 2024 · similarity 0.90)
- Optimize the Unseen - Fast NeRF Cleanup with Free Space Prior (NeurIPS 2025 · similarity 0.90)
- Bright-NeRF: Brightening Neural Radiance Field with Color Restoration from Low-Light RAW Images (AAAI 2025 · similarity 0.90)
- NeRF On-the-go: Exploiting Uncertainty for Distractor-free NeRFs in the Wild (CVPR 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.