Research map / Detection and segmentation / 3D detection and driving
Towards Generalization Across Depth for Monocular 3D Object Detection
Andrea Simonelli, Samuel Rota Buló, Lorenzo Porzi, Elisa Ricci, Peter Kontschieder
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 virtual views to MonoDIS (ablation) | MonoDIS · KITTI3D val, Car 3D | AP Moderate: 7.60 -> 10.89 | Improved |
| add bin-based distance estimation (5m bins) to MonoDIS (ablation) | MonoDIS · KITTI3D val, Car 3D | AP Moderate: 7.60 -> 5.34 | Hurt |
| add bin-based distance estimation (10m bins) to MonoDIS (ablation) | MonoDIS · KITTI3D val, Car 3D | AP Moderate: 7.60 -> 8.36 | Improved |
| use virtual views instead of bin-based estimation in MonoDIS (ablation) | MonoDIS · KITTI3D val, Car 3D | AP Moderate: 8.36 -> 10.89 | Improved |
| use virtual views instead of bin-based estimation (Cyclist) (ablation) | MonoDIS · KITTI3D val, Cyclist 3D | AP Moderate: 1.80 -> 1.07 | Hurt |
| reduce depth resolution Zres from 10m to 5m (ablation) | MoVi-3D · KITTI3D val, Car 3D | AP Moderate: 9.54 -> 11.13 | Improved |
| reduce depth resolution Zres from 20m to 10m (ablation) | MoVi-3D · KITTI3D val, Car 3D | AP Moderate: 6.18 -> 9.54 | Improved |
| reduce depth resolution Zres from 10m to 5m (ablation) | MoVi-3D · KITTI3D val, Pedestrian 3D | AP Moderate: 1.27 -> 5.52 | Improved |
| reduce depth resolution Zres from 20m to 10m (ablation) | MoVi-3D · KITTI3D val, Cyclist 3D | AP Moderate: 0.10 -> 0.51 | Improved |
| add virtual views (MoVi-3D vs MonoDIS), train far / val near (ablation) | MonoDIS · KITTI3D Car, train far / val near | 3D AP Moderate: 0.1 -> 1.9 | Improved |
| add virtual views (MoVi-3D vs MonoDIS), train near / val far (ablation) | MonoDIS · KITTI3D Car, train near / val far | 3D AP Moderate: 0.1 -> 1.4 | Improved |
| add virtual views (MoVi-3D vs MonoDIS), train near+far / val middle (ablation) | MonoDIS · KITTI3D Car, train near+far / val middle | 3D AP Moderate: 0.4 -> 10.6 | 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.
- M3D-RPN — M3D-RPN: Monocular 3D Region Proposal Network for Object Detection (ICCV 2019) · better on 6 of 7 shared results
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
- MonoDistill: Learning Spatial Features for Monocular 3D Object Detection (ICLR 2022 · similarity 0.90)
- MonoDGP: Monocular 3D Object Detection with Decoupled-Query and Geometry-Error Priors (CVPR 2025 · similarity 0.90)
- Rethinking Geometric Depth in Monocular 3D Object Detection: A Projection-Consistent Reformulation (NeurIPS 2026 · similarity 0.90)
- LeAD-M3D: Leveraging Asymmetric Distillation for Real-Time Monocular 3D Detection (ECCV 2026 · similarity 0.90)
- MonoMAE: Enhancing Monocular 3D Detection through Depth-Aware Masked Autoencoders (NeurIPS 2024 · similarity 0.89)
- MoGDE: Boosting Mobile Monocular 3D Object Detection with Ground Depth Estimation (NeurIPS 2022 · 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.
