Research map / 3D vision / Face and shape
LNEM: Lunar Neural Elevation Model
Suwan Lee, Jo Ryeong Yim, Kibaek Park, Dong-Gyu Kim, Eunhyeuk Kim, Minsup Jeong, Chae Kyung Sim, Seokju Lee
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 shadow modeling (sun visibility and indirect illumination branches) (ablation) | LNEM · LROC NAC, 8 regions | RMSELOLA: Apollo 15: 10.602 -> 8.602; lower in 7 of 8 cases | Improved |
| add shadow modeling (ablation) | LNEM · LROC NAC Apollo 16 (two views) | RMSELOLA: 1.970 -> 4.630 | Hurt |
| add shadow modeling (ablation) | LNEM · 8 NAC regions | sigma2_total: Lacus Mortis Pit: 17.8351 -> 6.2285 | Improved |
| replace Fourier positional encoding with multi-resolution hash encoding (L=32) (ablation) | LNEM · NAC Apollo 15 PSNR | PSNR: 29.01 -> 48.41 | Improved |
| replace Fourier positional encoding with hash encoding (ablation) | LNEM · NAC Apollo 17 PSNR | PSNR: 33.38 -> 48.39 | Improved |
| vary fraction of rays with depth supervision (10%, 40%, 100%) (ablation) | LNEM · Apollo 17 | RMSEcorr: 2.444, 2.098, 1.864 m | No clear effect |
| use rigorous pushbroom sensor model instead of RPC approximation | LNEM vs Sat-NeRF · LROC NAC, 8 regions | RMSEcorr: Apollo 15: 39.196 -> 1.565 | Improved |
| add depth supervision | Sat-NeRF/LNEM vs EO-NeRF · LROC NAC | vertical offset | Improved |
| use poorer LUTI reconstructed SPICE kernels instead of smithed NAC kernels | LNEM · KPLO LUTI Apollo 15 | RMSEcorr: 29.071 | Hurt |
| use three input views instead of two | LNEM · LROC NAC | RMSEcorr: 0.67 to 3.69 m vs 4.10 to 5.67 m | Improved |
Closest papers by meaning
- Geometric Foundation Model Distillation for Efficient Lunar 3D Reconstruction (ECCV 2026 · similarity 0.87)
- Mars-Bench: A Benchmark for Evaluating Foundation Models for Mars Science Tasks (NeurIPS 2025 · similarity 0.82)
- Seeing Across Skies and Streets: Feedforward 3D Reconstruction from Satellite, Drone, and Ground Images (NeurIPS 2026 · similarity 0.81)
- Dual-branch Graph Feature Learning for NLOS Imaging (AAAI 2025 · similarity 0.79)
- LRM: Large Reconstruction Model for Single Image to 3D (ICLR 2024 · similarity 0.79)
- Martian World Model: Controllable Video Synthesis with Physically Accurate 3D Reconstructions (NeurIPS 2025 · similarity 0.79)
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.
