Research map / Detection and segmentation / 3D detection and driving
Remove Projective LiDAR Depthmap Artifacts via Exploiting Epipolar Geometry
Shengjie Zhu, Girish Chandar Ganesan, Abhinav Kumar, Xiaoming Liu
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 |
|---|---|---|---|
| apply RePLAy artifact removal instead of the modified half-occlusion baseline to LiDAR depthmaps | LiDAR projected depthmap · KITTI Stereo GT (foreground) | delta_0.5: 0.919 -> 0.925 | Improved |
| add auto-calibration of the virtual LiDAR camera origin (ablation) | RePLAy · KITTI Eigen semi-dense GT (all) | SIlog: 5.680 -> 4.592 | Improved |
| train ZoeDepth on RePLAy-cleaned depthmaps instead of raw | ZoeDepth · KITTI Eigen (from scratch) | SIlog: 13.501 -> 8.674 | Improved |
| train ZoeDepth on RePLAy depthmaps instead of raw on datasets without stereo | ZoeDepth · nuScenes Val (foreground) | SIlog: 6.339 -> 4.905 | Improved |
| train ZoeDepth on RePLAy depthmaps instead of raw | ZoeDepth · Waymo Val (foreground) | SIlog: 9.116 -> 7.322 | Improved |
| use RePLAy depthmaps instead of raw, across rain, night, sunny conditions | ZoeDepth · nuScenes Val by condition | SIlog: Night 7.481 -> 5.374 | Improved |
| train ZoeDepth on RePLAy instead of semi-dense GT | ZoeDepth · KITTI Stereo (all) | SIlog: 7.971 -> 8.588 | Mixed |
| retrain CaDDN with RePLAy-cleaned depthmaps | CaDDN · KITTI Val Car AP3D | AP3D: Easy 21.60 -> 22.88; Hard 13.24 -> 13.18 | Mixed |
| retrain MonoDTR with RePLAy-cleaned depthmaps | MonoDTR · KITTI Val Car AP3D | AP3D: Easy 24.50 -> 26.15; Mod 18.66 -> 18.77; Hard 15.69 -> 15.69 | Improved |
Closest papers by meaning
- Multimodal Virtual Point 3D Detection (NeurIPS 2021 · similarity 0.87)
- Look Before You Fuse: 2D-Guided Cross-Modal Alignment for Robust 3D Detection (CVPR 2026 · similarity 0.87)
- Pseudo-LiDAR++: Accurate Depth for 3D Object Detection in Autonomous Driving (ICLR 2020 · similarity 0.87)
- Flexible Depth Completion for Sparse and Varying Point Densities (CVPR 2024 · similarity 0.86)
- Blurred LiDAR for Sharper 3D: Robust Handheld 3D Scanning with Diffuse LiDAR and RGB (CVPR 2025 · similarity 0.85)
- Mind The Edge: Refining Depth Edges in Sparsely-Supervised Monocular Depth Estimation (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.
