Research map / 3D vision / 3D generation
OmniHD-Scenes: A Next-Generation Multimodal Dataset for Autonomous Driving
Lianqing Zheng, Long Yang, Qunshu Lin, Wenjin Ai, Minghao Liu, Shouyi Lu, Jianan Liu, Hongze Ren, Jingyue Mo, Xiaokai Bai, Jie Bai, Zhixiong Ma and 1 more
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 |
|---|---|---|---|
| downsample LiDAR from 128 to 32 beams (ablation) | PointPillars · 3D detection | mAP/ODS: -3.91 mAP, -2.88 ODS | Hurt |
| replace PointPillars with RadarPillarNet for 4D radar | 4D radar detector · 3D detection | mAP/ODS: +1.06 mAP, +0.60 ODS | Improved |
| add temporal information to BEVFormer (ablation) | BEVFormer R50 544x960 · 3D detection | mAP/ODS: +2.68 mAP, +2.44 ODS | Improved |
| increase resolution to 864x1536 and backbone to R101-DCN (ablation) | BEVFormer · 3D detection | mAP: 26.49 -> 30.10 | Improved |
| evaluate under night and rainy conditions | vision-based detectors · 3D detection | mAP: -1.23 mAP to -1.99 mAP | Hurt |
| evaluate under night and rainy conditions | 4D radar and radar/camera fusion detectors · 3D detection | mAP: +1.48 to +4.90 mAP | Improved |
| use 4D radar instead of LiDAR for velocity | RadarPillarNet · 3D detection | mAVE: 0.6982 vs. 1.8731 | Improved |
| add temporal modelling to BEVFormer (ablation) | BEVFormer R50 occupancy · 3D occupancy | mIoU: +1.26 mIoU | Improved |
| increase resolution to 864x1536 and backbone to R101-DCN (ablation) | BEVFormer-T occupancy · 3D occupancy | mIoU: 16.23 -> 17.49 | Improved |
| fuse 4D radar with camera | camera occupancy · 3D occupancy | mIoU: 17.49 vs. 16.24 mIoU | Mixed |
| adverse conditions on camera-only occupancy | camera-only occupancy · 3D occupancy | mIoU: 1.1-1.65 mIoU decline | Hurt |
| adverse conditions on 4D radar and camera fusion occupancy | BEVFusion and M-CONet · 3D occupancy | mIoU: 0.78-0.88 mIoU decrease | Hurt |
Closest papers by meaning
- One Million Scenes for Autonomous Driving: ONCE Dataset (NeurIPS 2021 · similarity 0.92)
- The Road Ahead in Autonomous Driving: The KITScenes Multimodal Dataset (NeurIPS 2026 · similarity 0.90)
- UniScene: Unified Occupancy-centric Driving Scene Generation (CVPR 2025 · similarity 0.89)
- OpenAD: Open-World Autonomous Driving Benchmark for 3D Object Detection (NeurIPS 2025 · similarity 0.88)
- MAN TruckScenes: A multimodal dataset for autonomous trucking in diverse conditions (NeurIPS 2024 · similarity 0.88)
- OLiDM: Object-aware LiDAR Diffusion Models for Autonomous Driving (AAAI 2025 · 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.
