Research map / Detection and segmentation / 3D detection and driving
Monocular 3D Object Detection with Depth from Motion
Tai Wang, Jiangmiao Pang, Dahua Lin
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 flip augmentation (ablation) | DfM stereo cost volume baseline · KITTI val car | AP3D Moderate: 12.93 -> 13.92 | Improved |
| add rescale augmentation (ablation) | DfM stereo cost volume baseline · KITTI val car | AP3D Moderate: 13.92 -> 15.32 | Improved |
| add monocular compensation to stereo-only depth (ablation) | DfM · KITTI val car | AP3D Moderate: 15.32 -> 18.82 | Improved |
| add stereo branch to mono-only (ablation) | DfM · KITTI val car | AP3D Moderate: 15.30 -> 18.82 | Improved |
| add stereo to mono at hard level (ablation) | DfM · KITTI val car | AP3D Hard: 14.05 -> 16.47 | Mixed |
| use later preceding frame (1st to 3rd) at inference (ablation) | DfM · KITTI val car | AP3D Moderate: 17.27 -> 17.62 -> 17.96 | Improved |
| use quaternion rotation instead of Euler (ablation) | pose-free DfM · KITTI val car | AP3D Moderate: 15.03 -> 16.93 | Improved |
| add reprojection photometric supervision (ablation) | pose-free DfM · KITTI val car | AP3D Moderate: 16.93 -> 18.49 | Improved |
| use group-wise fusion instead of shared-weight monocular branch (ablation) | DfM · KITTI val car | AP3D Moderate: 15.99 -> 16.52 | Improved |
| use point-wise fusion instead of group-wise fusion (ablation) | DfM · KITTI val car | AP3D Moderate: 16.52 -> 18.82 | Improved |
| replace CE with focal loss (gamma=2) for depth (ablation) | DfM · KITTI val car | AP3D Moderate: 18.82 -> 18.76 | No clear effect |
| add foreground-balanced weights to depth loss (ablation) | DfM · KITTI val car | AP3D Moderate: 18.82 -> 19.11 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- MonoPair — MonoPair: Monocular 3D Object Detection Using Pairwise Spatial Relationships (CVPR 2020) · 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 0.90)
- Kinematic 3D Object Detection in Monocular Video (ECCV 2020 · similarity 0.90)
- MonoDGP: Monocular 3D Object Detection with Decoupled-Query and Geometry-Error Priors (CVPR 2025 · similarity 0.90)
- MonoCD: Monocular 3D Object Detection with Complementary Depths (CVPR 2024 · similarity 0.89)
- Depth2Pose: A Pose-Based Benchmark for Monocular Depth Estimation without Ground-Truth Depth (NeurIPS 2026 · similarity 0.89)
- MonoDiff: Monocular 3D Object Detection and Pose Estimation with Diffusion Models (CVPR 2024 · similarity 0.89)
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.
