Research map / Point clouds / Segmentation and detection
RESOLVE: A Multi-Resolution and Multi-Modal Dataset for Roadside Cooperative Perception
Shaozu Ding, Linan Song, Marco Vincenzi, Dajiang Suo
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 |
|---|---|---|---|
| increase LiDAR resolution from low to medium (ablation) | all 3D detectors · RESOLVE 3D detection | mAP: about 7.4% | Improved |
| increase LiDAR resolution from medium to high (ablation) | all 3D detectors · RESOLVE 3D detection | mAP: about 1.0% | No clear effect |
| increase LiDAR resolution from low to medium for LION (ablation) | LION · RESOLVE 3D detection | mAP: 88.1 -> 95.1 | Improved |
| increase LiDAR resolution from low to medium for Mamba models (ablation) | Mamba detectors · RESOLVE 3D detection | mAP: about 9.3% | Improved |
| add camera fusion to a LiDAR-only sparse-conv detector at low resolution | TransFusion-L · RESOLVE 3D detection (low-res LiDAR) | mAP: 82.5 -> 86.3 | Improved |
| add camera fusion to a LiDAR-only sparse-conv detector at medium resolution | TransFusion-L · RESOLVE 3D detection (mid-res LiDAR) | mAP: 86.6 -> 92.6 | Improved |
| let multimodal loss gradient update the LiDAR backbone (ablation) | BEVFusion · RESOLVE 3D detection (high-res LiDAR) | mAP: 89.4 -> 93.1 | Improved |
| train and test with matched LiDAR resolution rather than low-res training transferred to higher resolution (ablation) | Voxel Mamba · RESOLVE 3D detection (resolution mismatch) | mAP: 40% drop | Improved |
| test high-res trained model on medium resolution (ablation) | Voxel Mamba · RESOLVE 3D detection (resolution mismatch) | mAP: 4.6% drop | No clear effect |
| increase LiDAR resolution from low to medium (ablation) | tracking methods · RESOLVE tracking | AMOTP: 5.1% | Improved |
| increase LiDAR resolution from medium to high (ablation) | tracking methods · RESOLVE tracking | AMOTP: 14.1% | Improved |
| increase LiDAR resolution (ablation) | tracking methods · RESOLVE tracking | AMOTA | Mixed |
Closest papers by meaning
- Effective Multi-sensor Conditioning for Street-view Novel-view Synthesis (NeurIPS 2026 · similarity 0.90)
- RCooper: A Real-world Large-scale Dataset for Roadside Cooperative Perception (CVPR 2024 · similarity 0.89)
- UrbanIng-V2X: A Large-Scale Multi-Vehicle, Multi-Infrastructure Dataset Across Multiple Intersections for Cooperative Perception (NeurIPS 2025 · similarity 0.89)
- TUMTraf V2X Cooperative Perception Dataset (CVPR 2024 · similarity 0.88)
- HoloVIC: Large-scale Dataset and Benchmark for Multi-Sensor Holographic Intersection and Vehicle-Infrastructure Cooperative (CVPR 2024 · similarity 0.87)
- LidaRF: Delving into Lidar for Neural Radiance Field on Street Scenes (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.
