Research map / Detection and segmentation / 3D detection and driving
Train in Germany, Test in the USA: Making 3D Object Detectors Generalize
Yan Wang, Xiangyu Chen, Yurong You, Li Erran Li, Bharath Hariharan, Mark Campbell, Kilian Q. Weinberger, Wei-Lun Chao
Check 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 |
|---|---|---|---|
| Train on KITTI and test on Waymo instead of training and testing on Waymo | PointRCNN · Waymo validation, Easy | AP_3D: 85.3 -> 11.9 | Hurt |
| Normalize source training boxes and point clouds to target mean car sizes | PointRCNN trained on KITTI · Waymo validation, Easy | AP_3D: 11.9 -> 53.3 | Improved |
| Fine-tune with 10 labeled target instances | PointRCNN trained on KITTI · Waymo validation, Easy | AP_3D: 11.9 -> 70.9 | Improved |
| Directly add target-source mean size differences to predicted box sizes | PointRCNN trained on KITTI · Waymo validation, Easy | AP_3D: 11.9 -> 16.2 | Improved |
| Directly adjust predicted box sizes by the mean size difference rather than fine-tuning normalized training data | PointRCNN cross-dataset adaptation · Cross-dataset validation | AP_3D: Waymo: 11.9 -> 16.2; Argoverse: 22.3 -> 7.9 | Mixed |
| Replace detected box sizes with ground-truth sizes while keeping centers and rotations (ablation) | PointRCNN cross-dataset detections · Waymo validation, Easy | AP_3D: +65.2 | Improved |
| Apply point resampling before PIXOR (ablation) | PIXOR trained on Waymo · Cross-dataset detection | AP | Improved |
Closest papers by meaning
- Unsupervised Adaptation from Repeated Traversals for Autonomous Driving (NeurIPS 2022 · similarity 0.90)
- Pseudo-LiDAR From Visual Depth Estimation: Bridging the Gap in 3D Object Detection for Autonomous Driving (CVPR 2019 · similarity 0.90)
- One Million Scenes for Autonomous Driving: ONCE Dataset (NeurIPS 2021 · similarity 0.89)
- LaserNet: An Efficient Probabilistic 3D Object Detector for Autonomous Driving (CVPR 2019 · similarity 0.89)
- An Empirical Study of the Generalization Ability of Lidar 3D Object Detectors to Unseen Domains (CVPR 2024 · similarity 0.89)
- SAFDNet: A Simple and Effective Network for Fully Sparse 3D Object Detection (CVPR 2024 · 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.
