Research map / 3D vision / Depth and stereo
Deep Non-Line-of-Sight Reconstruction
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 a densely connected mechanism to the regressor (ablation) | Deep NLoS Reconstruction · NLoS reconstruction | reconstruction error: about 10% | Helped |
| Add Poisson noise and bias sensor augmentation (ablation) | ShapeNet-trained model · noisy synthetic inputs | reconstruction quality | Helped |
| Add Poisson noise and bias sensor augmentation (ablation) | ShapeNet-trained model · DiffuseS real-world input | reconstruction quality | Helped |
| Train on FlatNet instead of Redwood depth maps | Deep NLoS Reconstruction · synthetic test cases | accuracy: 0.82 vs 0.79; 0.99 vs 0.67; 0.94 vs 0.69 | Mixed |
| Train on Redwood instead of ShapeNet depth maps | Deep NLoS Reconstruction · synthetic test cases | RMSE: 0.23 vs 0.21; 0.04 vs 0.20; 0.19 vs 0.32 | Mixed |
| Downsample spatial input resolution by 2×2 (ablation) | ShapeNet-trained model · synthetic and real-world inputs | reconstruction quality | Hurt |
| Downsample temporal input resolution by fourfold (ablation) | ShapeNet-trained model · synthetic and real-world inputs | reconstruction quality | No clear effect |
| Increase output resolution to 128×128 (ablation) | ShapeNet-trained model · real-world diffuse targets | reconstruction quality | Helped |
| Increase output resolution to 128×128 (ablation) | ShapeNet-trained model · real-world retroreflective targets | reconstruction quality | No clear effect |
| Increase Poisson noise in the transient input (ablation) | ShapeNet-trained model · synthetic test cases | reconstruction quality | Hurt |
Closest papers by meaning
- Neural Inverse Rendering from Propagating Light (CVPR 2025 · similarity 0.87)
- Real-time Acquisition and Reconstruction of Dynamic Volumes with Neural Structured Illumination (CVPR 2024 · similarity 0.87)
- Virtual Scanning: Unsupervised Non-line-of-sight Imaging from Irregularly Undersampled Transients (NeurIPS 2024 · similarity 0.86)
- UniSDF: Unifying Neural Representations for High-Fidelity 3D Reconstruction of Complex Scenes with Reflections (NeurIPS 2024 · similarity 0.86)
- Inverse Rendering using Multi-Bounce Path Tracing and Reservoir Sampling (ICLR 2025 · similarity 0.85)
- PlatoNeRF: 3D Reconstruction in Plato's Cave via Single-View Two-Bounce Lidar (CVPR 2024 · similarity 0.85)
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.