Research map / Detection and segmentation / 3D detection and driving
Densely Constrained Depth Estimator for Monocular 3D Object Detection
Yingyan Li, Yuntao Chen, Jiawei He, Zhaoxiang Zhang
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 dense edge-based depth candidates (DGDE) with 10 keypoints (ablation) | MonoFlex with GMW · KITTI Val cars (IoU 0.7) | AP3D|R40 Hard (%): 13.63 -> 14.93 | Improved |
| add DGDE to uncertainty weighting with 10 keypoints (ablation) | MonoFlex baseline · KITTI Val cars (IoU 0.7) | AP3D|R40 Moderate (%): 15.87 -> 16.09 | Improved |
| use 73 keypoints (1500 candidates) instead of 10 keypoints (ablation) | DGDE with GMW · KITTI Val cars (IoU 0.7) | AP3D|R40 Moderate (%): 16.91 -> 17.38 | Improved |
| replace uncertainty weighting by graph matching weighting (GMW) (ablation) | DGDE with 73 keypoints · KITTI Val cars (IoU 0.7) | AP3D|R40 Moderate (%): 16.53 -> 17.38 | Improved |
| add GMW to baseline without DGDE (ablation) | MonoFlex · KITTI Val cars (IoU 0.7) | AP3D|R40 Moderate (%): 15.87 -> 16.14 | Improved |
| start the regression loss at 50 epochs instead of epoch 0 (ablation) | GMW · KITTI Val cars (IoU 0.7) | AP3D|R40 Moderate (%): 16.83 -> 17.38 | Improved |
| add regression loss starting at 50 epochs to classification loss (ablation) | GMW · KITTI Val cars (IoU 0.7) | AP3D|R40 Moderate (%): 17.03 -> 17.38 | Improved |
| use 1500 candidate edges instead of 50 (ablation) | DCD · KITTI Val cars (IoU 0.7) | AP3D|R40 Moderate (%): 16.93 -> 17.38 | Improved |
| use 1500 candidate edges instead of all 2628 (ablation) | DCD · KITTI Val cars (IoU 0.7) | AP3D|R40 Moderate (%): 16.98 -> 17.38 | Improved |
| use DCD depth estimation and weighting instead of AutoShape's (ablation) | MonoFlex baseline · KITTI Val cars (IoU 0.7) | AP3D|R40 Moderate (%): 16.48 -> 17.38 | Improved |
| replace MonoFlex with DCD | MonoFlex · KITTI Test cars | AP3D|R40 Moderate (%): 13.89 -> 15.90 | Improved |
| replace MonoFlex with DCD | MonoFlex · Waymo Val LEVEL_1 IoU 0.7 | 3D AP (%): 11.70 -> 12.57 | Improved |
Closest papers by meaning
- MonoCD: Monocular 3D Object Detection with Complementary Depths (CVPR 2024 · similarity 0.90)
- Objects Are Different: Flexible Monocular 3D Object Detection (CVPR 2021 · similarity 0.89)
- Single View Stereo Matching (CVPR 2018 · similarity 0.88)
- Diversity Matters: Fully Exploiting Depth Clues for Reliable Monocular 3D Object Detection (CVPR 2022 · similarity 0.88)
- MonoDGP: Monocular 3D Object Detection with Decoupled-Query and Geometry-Error Priors (CVPR 2025 · similarity 0.88)
- MonoDiff: Monocular 3D Object Detection and Pose Estimation with Diffusion Models (CVPR 2024 · similarity 0.87)
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.
