Research map / Detection and segmentation / 3D detection and driving
SPAN: Spatial-Projection Alignment for Monocular 3D Object Detection
Yifan Wang, Yian Zhao, Fanqi Pu, Xiaochen Yang, Yang Tang, Xi Chen, Wenming Yang
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 Spatial Point Alignment loss alone, without HTL (ablation) | MonoDGP · KITTI val Car | AP3D R40 Mod.: 22.34 -> 21.92 | Hurt |
| add 3D-2D Projection Alignment loss alone, without HTL (ablation) | MonoDGP · KITTI val Car | AP3D R40 Mod.: 22.34 -> 21.80 | Hurt |
| add Hierarchical Task Learning alone (ablation) | MonoDGP · KITTI val Car | AP3D R40 Mod.: 22.34 -> 22.56 | Improved |
| add HTL schedule to both SPAN losses (ablation) | MonoDGP + SPAN · KITTI val Car | AP3D R40 Mod.: 21.85 -> 23.26 | Improved |
| use HTL instead of linear weight schedule (ablation) | MonoDGP + SPAN · KITTI val Car | AP3D R40 Mod.: 22.95 -> 23.26 | Improved |
| use MGIoU instead of L1 for corner loss (ablation) | MonoDGP + HTL · KITTI val Car | AP3D R40 Mod.: 22.68 -> 22.89 | Improved |
| use MGIoU instead of exact 3D IoU for corner loss (ablation) | MonoDGP + HTL · KITTI val Car | AP3D R40 Mod.: 22.75 -> 22.89 | Improved |
| add Spatial Point Alignment to projection loss with HTL (ablation) | MonoDGP · KITTI val Car | AP3D R40 Mod.: 22.97 -> 23.26 | Improved |
| add 3D-2D Projection Alignment to corner loss with HTL (ablation) | MonoDGP · KITTI val Car | AP3D R40 Mod.: 22.89 -> 23.26 | Improved |
| raise loss weights from 0.5 to 1.0 (ablation) | MonoDGP + SPAN · KITTI val Car | AP3D R40 Mod.: 22.81 -> 23.26 | Improved |
| raise loss weights from 1.0 to 2.0 (ablation) | MonoDGP + SPAN · KITTI val Car | AP3D R40 Mod.: 23.26 -> 22.66 | Hurt |
| add 15px uniform noise to 2D boxes during training (ablation) | MonoDGP + SPAN · KITTI val Car | AP3D R40 Mod.: 23.26 -> 13.51 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- MonoDGP — MonoDGP: Monocular 3D Object Detection with Decoupled-Query and Geometry-Error Priors (CVPR 2025) · better on 12 of 12 shared results
- MonoDETR — MonoDETR: Depth-guided Transformer for Monocular 3D Object Detection (ICCV 2023) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Rethinking Geometric Depth in Monocular 3D Object Detection: A Projection-Consistent Reformulation (NeurIPS 2026 · similarity 0.89)
- Weakly Supervised Monocular 3D Object Detection by Spatial-Temporal View Consistency (TPAMI 2024 · similarity 0.88)
- Exploring Geometric Consistency for Monocular 3D Object Detection (CVPR 2022 · similarity 0.88)
- MonoPlace3D: Learning 3D-Aware Object Placement for 3D Monocular Detection (CVPR 2025 · similarity 0.88)
- Monocular 3D Object Detection with Bounding Box Denoising in 3D by Perceiver (ICCV 2023 · similarity 0.88)
- MonoDGP: Monocular 3D Object Detection with Decoupled-Query and Geometry-Error Priors (CVPR 2025 · 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.
