Research map / 3D vision / Depth and stereo
3D Scene Inference from Transient Histograms
Sacha Jungerman, Atul Ingle, Yin Li, Mohit Gupta
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 learned transient depth network instead of bilinear upsampling of peak depths | Ours · NYUv2 (4x3 grid) | delta<1.05: 0.285 -> 0.335 | Improved |
| use learned transient depth network instead of bilinear upsampling of peak depths | Ours · NYUv2 (20x15 grid) | delta<1.05: 0.583 -> 0.624 | Improved |
| use transient network instead of super-resolution baseline | Ours · NYUv2 (4x3 grid) | delta<1.05: 0.340 -> 0.335 | No clear effect |
| use transient network instead of super-resolution baseline | Ours · NYUv2 (4x3 grid) | AbsRel: 0.147 -> 0.126 | Improved |
| use transient network instead of bilinear | Ours · NYUv2 (20x15 grid) | Log10: 0.038 -> 0.060 | Hurt |
| add RGB-guided depth refinement (FDKN+DPT) | Ours · NYUv2 (20x15 grid) | delta<1.05: 0.624 -> 0.707 | Improved |
| add RGB-guided depth refinement (FDKN+DPT) | Ours · NYUv2 (20x15 grid) | AbsRel: 0.073 -> 0.053 | Improved |
| increase transient grid from 4x3 to 20x15 | Ours · NYUv2 | delta<1.05: 0.335 -> 0.624 | Improved |
| use analysis-by-synthesis instead of theoretical estimator for plane angle | AbS · synthetic planes | theta_n error: 87% of the time | Improved |
| use 4x3 SPAD transients instead of RGB monocular depth network | Ours 4x3 · NYUv2 | delta<1.25^3: 0.998 -> 0.981 | No clear effect |
| use 4x3 transients instead of DORN in sensing cost | Ours 4x3 · NYUv2 | inference time: 112.0 ms -> 2.77 ms | Improved |
Closest papers by meaning
- Reconstructing Transient Images From Single-Photon Sensors (CVPR 2017 · similarity 0.89)
- Deep Non-Line-of-Sight Reconstruction (CVPR 2020 · similarity 0.87)
- Recovering Parametric Scenes from Very Few Time-of-Flight Pixels (ICCV 2025 · similarity 0.87)
- Towards 3D Vision with Low-Cost Single-Photon Cameras (CVPR 2024 · similarity 0.86)
- Event Fields: Capturing Light Fields at High Speed, Resolution, and Dynamic Range (CVPR 2025 · similarity 0.86)
- Neural Inverse Rendering from Propagating Light (CVPR 2025 · similarity 0.86)
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.
