Research map / Detection and segmentation / 3D detection and driving
DID-M3D: Decoupling Instance Depth for Monocular 3D Object Detection
Liang Peng, Xiaopei Wu, Zheng Yang, Haifeng Liu, Deng Cai
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 |
|---|---|---|---|
| decouple instance depth into visual and attribute depth (ablation) | grid baseline · KITTI val Car moderate | AP3D R40: 11.24 -> 13.24 | Improved |
| decouple instance depth on the naive baseline (ablation) | naive baseline · KITTI val Car moderate | AP_BEV/AP3D: 13.43/8.70 -> 16.49/10.94 | Improved |
| add both depth uncertainties (ablation) | DID-M3D decoupled depth · KITTI val Car moderate | AP3D R40: 13.24 -> 13.91 | Improved |
| add instance depth confidence to 2D confidence (ablation) | DID-M3D · KITTI val Car moderate | AP3D R40: 13.91 -> 15.57 | Improved |
| add adaptive aggregation over grids (ablation) | DID-M3D · KITTI val Car moderate | AP3D R40: 15.57 -> 16.12 | Improved |
| add adaptive aggregation to coupled baseline (ablation) | coupled baseline · KITTI val Car moderate | AP_BEV/AP3D: 0.1/0.03 gains | No clear effect |
| add affine augmentation with scaled visual depth (ablation) | DID-M3D · KITTI val Car moderate | AP3D R40: 12.76 -> 16.12 | Improved |
| add affine augmentation but scale attribute depth (ablation) | DID-M3D · KITTI val Car moderate | AP3D R40: 12.76 -> 12.65 | Hurt |
| use grid 1x1 instead of 7x7 (ablation) | DID-M3D · KITTI val Car moderate | AP3D R40: 16.12 -> 13.01 | Hurt |
| use grid 19x19 instead of 7x7 (ablation) | DID-M3D · KITTI val Car moderate | AP3D R40: 16.12 -> 15.13 | Hurt |
| add grid design to naive baseline (ablation) | naive baseline · KITTI val Car moderate | AP_BEV/AP3D: 1.0/0.24 gains | Improved |
| add DID-M3D components to baseline | baseline detector · KITTI val Pedestrian moderate | AP3D R40: 3.88 -> 5.87 | Improved |
Later papers that compared with this method
5 later accepted papers report DID-M3D as a baseline in their results tables; 4 of them report a better number on most of the results they share with it.
- Object-Scene-Camera Decomposition and Recomposition for Data Efficient Monocular 3D Object Detection (IJCV 2026 · better on 4 of 4 shared results)
- Improving Distant 3D Object Detection Using 2D Box Supervision (CVPR 2024 · better on 1 of 2 shared results)
- Learning Occupancy for Monocular 3D Object Detection (CVPR 2024 · better on 1 of 1 shared result)
- MonoNeRD: NeRF-like Representations for Monocular 3D Object Detection (ICCV 2023 · better on 3 of 4 shared results)
- Depth-discriminative Metric Learning for Monocular 3D Object Detection (NeurIPS 2023 · better on 3 of 3 shared results)
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.
- MonoJSG — MonoJSG: Joint Semantic and Geometric Cost Volume for Monocular 3D Object Detection (CVPR 2022) · 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
- Objects Are Different: Flexible Monocular 3D Object Detection (CVPR 2021 · similarity 0.92)
- Diversity Matters: Fully Exploiting Depth Clues for Reliable Monocular 3D Object Detection (CVPR 2022 · similarity 0.91)
- MonoCD: Monocular 3D Object Detection with Complementary Depths (CVPR 2024 · similarity 0.91)
- MonoDGP: Monocular 3D Object Detection with Decoupled-Query and Geometry-Error Priors (CVPR 2025 · similarity 0.89)
- MonoMAE: Enhancing Monocular 3D Detection through Depth-Aware Masked Autoencoders (NeurIPS 2024 · similarity 0.88)
- RefineAny3D: Depth Refinement as Semantic Alignment for Monocular 3D Detection (NeurIPS 2026 · similarity 0.88)
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.
