Research map / Detection and segmentation / 3D detection and driving
MonoDETR: Depth-guided Transformer for Monocular 3D Object Detection
Renrui Zhang, Han Qiu, Tai Wang, Ziyu Guo, Ziteng Cui, Yu Qiao, Hongsheng Li, Peng Gao
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 depth-guided transformer and depth predictor (ablation) | center-guided baseline (MonoDLE-like) · KITTI val (car, AP3D Mod.) | AP3D: 15.15 -> 20.61 | Improved |
| add depth cross-attention guidance to visual transformer (ablation) | MonoDETR · KITTI val (car, AP3D Mod.) | AP3D: 17.81 -> 20.61 | Improved |
| add transformer on top of depth predictor (ablation) | MonoDETR · KITTI val (car, AP3D Mod.) | AP3D: 16.05 -> 20.61 | Improved |
| use one-block global self-attention depth encoder (ablation) | MonoDETR · KITTI val (car, AP3D Mod.) | AP3D: 18.38 -> 20.61 | Improved |
| place depth cross-attention first in each decoder block (ablation) | depth-guided decoder · KITTI val (car, AP3D Mod.) | AP3D: 19.28 -> 20.61 | Improved |
| supervise foreground depth with object-wise labels instead of dense depth (ablation) | foreground depth map · KITTI val (car, AP3D Mod.) | AP3D: 19.85 -> 20.61 | Improved |
| use linear-increasing discretization instead of uniform or spacing-increasing (ablation) | foreground depth map · KITTI val (car, AP3D Mod.) | AP3D: 18.90 -> 20.61 | Improved |
| use learnable meter-wise depth positional encoding instead of sinusoidal or per-bin (ablation) | depth cross-attention · KITTI val (car, AP3D Mod.) | AP3D: 19.57 -> 20.61 | Improved |
| 2D sin/cos positional encoding instead of none (ablation) | depth cross-attention · KITTI val (car, AP3D Mod.) | AP3D: 18.94 -> 18.63 | Hurt |
| plug depth-guided modules into PETRv2 | PETRv2 · nuScenes val | NDS: 0.496 -> 0.508 | Improved |
| plug depth-guided modules into BEVFormer | BEVFormer · nuScenes val | NDS: 0.517 -> 0.526 | Improved |
Later papers that compared with this method
16 later accepted papers report MonoDETR as a baseline in their results tables; 15 of them report a better number on most of the results they share with it.
- Difficulty-Aware Label-Guided Denoising for Monocular 3D Object Detection (AAAI 2026 · better on 6 of 6 shared results)
- MonoCLUE: Object-Aware Clustering Enhances Monocular 3D Object Detection (AAAI 2026 · better on 6 of 6 shared results)
- Unleashing the Power of Chain-of-Prediction for Monocular 3D Object Detection (CVPR 2026 · better on 2 of 2 shared results)
- RARE: Learn to RAnk and REtrieve for Monocular 3D Object Detection (CVPR 2026 · better on 2 of 2 shared results)
- Object-Scene-Camera Decomposition and Recomposition for Data Efficient Monocular 3D Object Detection (IJCV 2026 · better on 2 of 2 shared results)
- SPAN: Spatial-Projection Alignment for Monocular 3D Object Detection (CVPR 2026 · better on 1 of 1 shared result)
- Towards Intrinsic-Aware Monocular 3D Object Detection (CVPR 2026 · better on 1 of 1 shared result)
- Test-Time Correction: An Online 3D Detection System via Visual Prompting (TPAMI 2025 · better on 6 of 6 shared results)
8 more not listed.
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- MonoDTR — MonoDTR: Monocular 3D Object Detection With Depth-Aware Transformer (CVPR 2022) · better on 3 of 3 shared results
- BEVFormer — BEVFormer: Learning Bird’s-Eye-View Representation from Multi-Camera Images via Spatiotemporal Transformers (ECCV 2022) · better on 1 of 1 shared result
- PETRv2 — PETR: Position Embedding Transformation for Multi-View 3D Object Detection (ECCV 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
- MonoDTR: Monocular 3D Object Detection With Depth-Aware Transformer (CVPR 2022 · similarity 0.94)
- MonoDGP: Monocular 3D Object Detection with Decoupled-Query and Geometry-Error Priors (CVPR 2025 · similarity 0.93)
- MonoMAE: Enhancing Monocular 3D Detection through Depth-Aware Masked Autoencoders (NeurIPS 2024 · similarity 0.92)
- MonoCD: Monocular 3D Object Detection with Complementary Depths (CVPR 2024 · similarity 0.91)
- MonoDiff: Monocular 3D Object Detection and Pose Estimation with Diffusion Models (CVPR 2024 · similarity 0.90)
- FD3D: Exploiting Foreground Depth Map for Feature-Supervised Monocular 3D Object Detection (AAAI 2024 · similarity 0.90)
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.
