Research map / Video / Video segmentation and tracking
Consistent Time-of-Flight Depth Denoising via Graph-Informed Geometric Attention
Weida Wang, Changyong He, Jin Zeng, Di Qiu
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 unrolled GLR module (single-frame) (ablation) | GIGA-ToF baseline network · DVToF (synthetic test) | MAE (m): 0.0409 -> 0.0357 | Improved |
| add unrolled GLR module (single-frame) (ablation) | GIGA-ToF baseline network · DVToF (synthetic test) | TEPE (m): 0.0793 -> 0.0734 | Improved |
| add depth-feature fusion across frames (ablation) | GIGA-ToF with unrolled GLR · DVToF (synthetic test) | MAE (m): 0.0357 -> 0.0238 | Improved |
| replace feature fusion with graph fusion (no attention) (ablation) | GIGA-ToF · DVToF (synthetic test) | MAE (m): 0.0238 -> 0.0219 | Improved |
| replace feature fusion with graph fusion (with attention) (ablation) | GIGA-ToF · DVToF (synthetic test) | TEPE (m): 0.0713 -> 0.0637 | Improved |
| add inter-frame attention to graph fusion (ablation) | GIGA-ToF · DVToF (synthetic test) | MAE (m): 0.0219 -> 0.0193 | Improved |
| add inter-frame attention to graph fusion (ablation) | GIGA-ToF · DVToF (synthetic test) | TEPE (m): 0.0702 -> 0.0637 | Improved |
| add inter-frame attention to feature fusion (ablation) | GIGA-ToF with feature fusion · DVToF (synthetic test) | MAE (m): 0.0238 -> 0.0214 | Improved |
| increase frame time step from 1 to 8 (ablation) | GIGA-ToF · DVToF (synthetic test) | MAE (m): 0.0190 -> 0.0210 | Hurt |
| increase frame time step from 1 to 2 (ablation) | GIGA-ToF · DVToF (synthetic test) | TEPE (m): 0.0634 -> 0.0608 | No clear effect |
| use multi-frame GIGA-ToF instead of best single-frame GLRUN | GIGA-ToF vs GLRUN · DVToF (synthetic test) | MAE (m): 0.0357 -> 0.0193 | Improved |
| use GIGA-ToF instead of WMF under augmented noise | GIGA-ToF vs WMF · DVToF with augmented noise | TEPE (m): 0.0950 -> 0.1102 | Hurt |
Closest papers by meaning
- Exploiting Dual-Correlation for Multi-frame Time-of-Flight Denoising (ECCV 2024 · similarity 0.90)
- Spatial Hierarchy Aware Residual Pyramid Network for Time-of-Flight Depth Denoising (ECCV 2020 · similarity 0.89)
- Deep End-to-End Time-of-Flight Imaging (CVPR 2018 · similarity 0.87)
- SVDC: Consistent Direct Time-of-Flight Video Depth Completion with Frequency Selective Fusion (CVPR 2025 · similarity 0.85)
- Injecting Frame-Event Complementary Fusion into Diffusion for Optical Flow in Challenging Scenes (NeurIPS 2025 · similarity 0.85)
- 3DMambaIPF: A State Space Model for Iterative Point Cloud Filtering via Differentiable Rendering (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.
