Research map / Video / Video segmentation and tracking
Exploiting Dual-Correlation for Multi-frame Time-of-Flight Denoising
Guanting Dong, Yueyi Zhang, Xiaoyan Sun, Zhiwei Xiong
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 dual-correlation estimation module (ablation) | MTDNet · HAMMER | overall MAE (cm): 0.122 -> 0.097 | Improved |
| add confidence map (CRRM) (ablation) | MTDNet · HAMMER | overall MAE (cm): 0.111 -> 0.097 | Improved |
| add residual channel attention block with residual pyramid combination (ablation) | MTDNet · HAMMER | overall MAE (cm): 0.313 -> 0.097 | Improved |
| add depth refinement module (ablation) | MTDNet · HAMMER | overall MAE (cm): 0.105 -> 0.097 | Improved |
| add inter-correlation (ablation) | MTDNet · HAMMER | overall MAE (cm), 4th quantile: 0.270 -> 0.207 | Improved |
| add intra-correlation (ablation) | MTDNet · HAMMER | overall MAE (cm), 1st quantile: 0.057 -> 0.050 | Improved |
| use dual-correlation instead of flow-based alignment (ablation) | MTDNet · HAMMER | overall MAE (cm): 0.114 -> 0.097 | Improved |
| use dual-correlation instead of deformable alignment (ablation) | MTDNet · HAMMER | overall MAE (cm): 0.108 -> 0.097 | Improved |
| vary number of intermediate frames between t and t+1 (3 best) (ablation) | MTDNet · TFT3D | overall MAE (cm): 0.88 (0 frame) -> 0.82 (3 frame) -> 0.86 (7 frame) | Mixed |
| use multi-frame MTDNet over single-frame SHARP-Net | MTDNet · TFT3D | overall MAE (cm): 1.02 -> 0.82 | Improved |
Closest papers by meaning
- Consistent Time-of-Flight Depth Denoising via Graph-Informed Geometric Attention (ICCV 2025 · similarity 0.90)
- Spatial Hierarchy Aware Residual Pyramid Network for Time-of-Flight Depth Denoising (ECCV 2020 · similarity 0.88)
- Efficient View Synthesis and 3D-Based Multi-Frame Denoising With Multiplane Feature Representations (CVPR 2023 · similarity 0.88)
- Deep End-to-End Time-of-Flight Imaging (CVPR 2018 · similarity 0.87)
- Physically-Grounded Turbulence Mitigation with Frame-Shared Degradation Parameters (CVPR 2026 · similarity 0.87)
- How I Warped Your Noise: a Temporally-Correlated Noise Prior for Diffusion Models (ICLR 2024 · 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.
