Research map / 3D vision / 3D generation
ClaraVid: A Holistic Scene Reconstruction Benchmark From Aerial Perspective With Delentropy-Based Complexity Profiling
Radu Beche, Sergiu Nedevschi
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 ClaraVid instead of Syndrone for training | DPT segmentation · UAVid | mIoU: 52.49 -> 57.50 | Improved |
| use ClaraVid instead of Syndrone for training | DPT segmentation · UDD | mIoU: 54.43 -> 58.22 | Improved |
| use ClaraVid instead of Syndrone for training | DPT segmentation · SkyScapes | mIoU: 42.75 -> 47.52 | Improved |
| use ClaraVid instead of Syndrone for training | DPT segmentation · Dronescapes | mIoU: 34.11 -> 37.30 | Improved |
| use ClaraVid instead of Syndrone for training | DPT segmentation · average synthetic-to-real gap | gap (%): 22.34 -> 17.82 | Improved |
| replace Nerfacto with Gaussian Splatting | reconstruction · ClaraVid NVS average | PSNR: 21.84 -> 23.67 | Improved |
| replace TensoRF with ZipNeRF | reconstruction · ClaraVid NVS average | SSIM: 0.4144 -> 0.5966 | Improved |
| train on 90H, test on 45L instead of matched 45L | Gaussian Splatting · ClaraVid viewpoint | PSNR: 22.98 -> 17.88 | Hurt |
| train on 90H, test on 45L instead of matched 45L (ablation) | Nerfacto · ClaraVid viewpoint | PSNR: 21.36 -> 18.23 | Hurt |
| train on 90H, test on 45L instead of matched 45L | Gaussian Splatting · ClaraVid viewpoint segmentation | mIoU: 56.53 -> 36.68 | Hurt |
| use delentropy instead of Shannon or GLCM entropy as complexity measure (ablation) | DSP · correlation with reconstruction error | Pearson correlation | Improved |
| higher mean delentropy of scene | reconstruction methods · ClaraVid NVS | PSNR: Rural #1 mu 4.18 PSNR 19.97 vs Urban #2 mu 3.29 PSNR 26.68 (Gaussian Splatting) | Hurt |
Closest papers by meaning
- AerialMegaDepth: Learning Aerial-Ground Reconstruction and View Synthesis (CVPR 2025 · similarity 0.86)
- Horizon-GS: Unified 3D Gaussian Splatting for Large-Scale Aerial-to-Ground Scenes (CVPR 2025 · similarity 0.85)
- UniSDF: Unifying Neural Representations for High-Fidelity 3D Reconstruction of Complex Scenes with Reflections (NeurIPS 2024 · similarity 0.84)
- DroneSplat: 3D Gaussian Splatting for Robust 3D Reconstruction from In-the-Wild Drone Imagery (CVPR 2025 · similarity 0.84)
- Multi-view Reconstruction via SfM-guided Monocular Depth Estimation (CVPR 2025 · similarity 0.84)
- Dual-branch Graph Feature Learning for NLOS Imaging (AAAI 2025 · similarity 0.84)
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.
