Research map / 3D vision / Depth and stereo
Photon-Starved Scene Inference Using Single Photon Cameras
Bhavya Goyal, 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 |
|---|---|---|---|
| add photon scale space images to training (ablation) | Vanilla Net (ResNet-18) · CUB-200-2011 S1 | top-1 accuracy: 21.35 -> 28.92 | Improved |
| add feature consistency (MSE) loss across PPP levels (ablation) | Vanilla Net w/ Photon Scaled Images · CUB-200-2011 S1 | top-1 accuracy: 28.92 -> 42.37 | Improved |
| use Photon Net instead of Curriculum Learning (ablation) | CUB-200-2011 classifier · CUB-200-2011 S1 | top-1 accuracy: 33.72 -> 42.37 | Improved |
| use Photon Net instead of N-step Student Teacher Learning (ablation) | CUB-200-2011 classifier · CUB-200-2011 S1 | top-1 accuracy: 35.79 -> 42.37 | Improved |
| BM3D denoising of input images (ablation) | Vanilla Net · CUB-200-2011 S1 | top-1 accuracy: 21.35 -> 25.52 | Improved |
| use Photon Net instead of Joint Denoising | classifier · CUB-200-2011 S1 | top-1 accuracy: 27.21 -> 42.37 | Improved |
| use Photon Net instead of Joint Denoising | classifier · CARS-196 S1 | top-1 accuracy: 34.51 -> 64.23 | Improved |
| use Photon Net instead of Student-Teacher Learning | classifier · CARS-196 S10 | top-1 accuracy: 73.51 -> 78.97 | Improved |
| use Photon Net instead of Joint Denoising | DenseDepth · NYUv2 S1 | rel: 0.209 -> 0.183 | Improved |
| use Photon Net instead of Joint Denoising | DenseDepth · NYUv2 S10 | delta1: 0.763 -> 0.793 | Improved |
| increase number of photon scale space levels from 2 to 5 (ablation) | Photon Net · CUB-200-2011 and CARS-196 S1 | top-1 accuracy | Improved |
| set MSE loss weight lambda to 25.0 (vs 0 to 50.0 range) (ablation) | Photon Net · CUB-200-2011 S1 | top-1 accuracy | Improved |
Closest papers by meaning
- Panoramas from Photons (ICCV 2023 · similarity 0.88)
- Predicting Important Photons for Energy-Efficient Single-Photon Videography (TPAMI 2025 · similarity 0.88)
- Inter-Photon-Limited Videography (CVPR 2026 · similarity 0.88)
- Towards 3D Vision with Low-Cost Single-Photon Cameras (CVPR 2024 · similarity 0.87)
- Generalized Event Cameras (CVPR 2024 · similarity 0.86)
- bit2bit: 1-bit quanta video reconstruction via self-supervised photon prediction (NeurIPS 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.
