Research map / Detection and segmentation / 3D detection and driving
LiDAR perception for driving
Build efficient LiDAR detectors, anomaly datasets, and generative models for driving perception.
75 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2019–2026) · an approach in 3D detection and driving, keyword cluster “bev · autonomous · radar”
State History Tried Checklist Preprints Who Export
Where it stands
Not enough benchmark data Fewer than four papers report the same benchmark, so there is no record to follow.
26 papers in 2023–24 and 32 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic grew, from 16.5% to 21.1%.
How this is decided
A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.
How it got here
Most cited paper of each year
- 2019 · LaserNet: An Efficient Probabilistic 3D Object Detector for Autonomous Driving (CVPR 2019 · 406 citations)
- 2020 · nuScenes: A Multimodal Dataset for Autonomous Driving (CVPR 2020 · 6,931 citations)
- 2021 · Multi-Modal Fusion Transformer for End-to-End Autonomous Driving (CVPR 2021 · 578 citations)
- 2022 · DAIR-V2X: A Large-Scale Dataset for Vehicle-Infrastructure Cooperative 3D Object Detection (CVPR 2022 · 560 citations)
- 2023 · V2V4Real: A Real-World Large-Scale Dataset for Vehicle-to-Vehicle Cooperative Perception (CVPR 2023 · 243 citations)
- 2024 · TUMTraf V2X Cooperative Perception Dataset (CVPR 2024 · 110 citations)
- 2025 · Language Prompt for Autonomous Driving (AAAI 2025 · 62 citations)
- 2026 · Griffin: Aerial-Ground Cooperative Detection and Tracking Dataset and Benchmark (AAAI 2026 · 3 citations)
Earliest papers here
- LaserNet: An Efficient Probabilistic 3D Object Detector for Autonomous Driving (CVPR 2019 · 406 citations)
- nuScenes: A Multimodal Dataset for Autonomous Driving (CVPR 2020 · 6,931 citations)
- Predicting Semantic Map Representations From Images Using Pyramid Occupancy Networks (CVPR 2020 · 260 citations)
What has been tried
57 papers here report 630 measured changes (a component added, removed or swapped, compared with a baseline): helped 74%, no clear effect 4%, hurt 21%, mixed 1%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| add cooperative vehicle-infrastructure camera-LiDAR fusion instead of vehicle-only camera-LiDAR perception | TUMTraf-V2X test set, south2 FOV · mAP3D | 76.40 -> 90.76; +14.36 3D mAP | Helped | TUMTraf V2X Cooperative Perception Dataset (CVPR 2024 · 110 citations) |
| Replace Average Fusion with CoBEVT fusion | DeepAccident validation · APA | 67.1 -> 69.5 | Helped | DeepAccident: A Motion and Accident Prediction Benchmark for V2X Autonomous Driving (AAAI 2024 · 97 citations) |
| jointly train detection and segmentation | nuScenes val · mIoU | 0.652 -> 0.656 | Helped | PanoOcc: Unified Occupancy Representation for Camera-based 3D Panoptic Segmentation (CVPR 2024 · 82 citations) |
| Remove prompt fusion and use all tracking instances as prompt predictions | NuPrompt validation · AMOTA | 0.259 -> 0.073 | Hurt | Language Prompt for Autonomous Driving (AAAI 2025 · 62 citations) |
| enable collaboration across CAVs | OPV2V semantic occupancy prediction · IoU | 38.52 -> 50.46 | Helped | Collaborative Semantic Occupancy Prediction with Hybrid Feature Fusion in Connected Automated Vehicles (CVPR 2024 · 36 citations) |
| Add data-space alignment | SemanticKITTI, nuScenes, and Waymo Open · mIoU | 54.03 -> 69.01; 59.84 -> 76.89; 65.39 -> 69.37 | Helped | Multi-Space Alignments Towards Universal LiDAR Segmentation (CVPR 2024 · 24 citations) |
| replace PFN with SPE | nuScenes validation · mAP | 48.72 -> 49.75 | Helped | SwiftPillars: High-Efficiency Pillar Encoder for Lidar-Based 3D Detection (AAAI 2024 · 12 citations) |
| Use continuous occupancy pre-training with balanced positive and negative query samples instead of depth rendering | Argoverse 2 Sensor · NFCD (m2) | 1.79 -> 0.83 | Helped | UnO: Unsupervised Occupancy Fields for Perception and Forecasting (CVPR 2024 · 12 citations) |
| Add multi-task detection and segmentation training to the M3Net occupancy model | nuScenes-Occupancy val · mIoU | 6% | Helped | M3Net: Multimodal Multi-task Learning for 3D Detection, Segmentation, and Occupancy Prediction in Autonomous Driving (AAAI 2025 · 9 citations) |
| Remove patch projection | aiMotive · all-point AP | -1.03 | Hurt | Hyperdimensional Uncertainty Quantification for Multimodal Uncertainty Fusion in Autonomous Vehicles Perception (CVPR 2025 · 9 citations) |
| Add pseudo point clouds to the convolutional feature branch | OpenAD 3D converter · AR | 18.38 -> 23.32 | Helped | OpenAD: Open-World Autonomous Driving Benchmark for 3D Object Detection (NeurIPS 2025 · 8 citations) |
| Add sparse-to-dense distillation | OPV2V average across corruptions · AP@0.5/0.7 | 79.75/70.54 -> 83.42/73.92 | Helped | DSRC: Learning Density-Insensitive and Semantic-Aware Collaborative Representation Against Corruptions (AAAI 2025 · 7 citations) |
What reviewers will expect
Datasets and baselines named in the 17 papers of this approach from 2025–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Datasets
- nuScenes · 10 of 17 papers
- KITTI · 3 of 17 papers
- OPV2V · 2 of 17 papers
- DAIR-V2X · 2 of 17 papers
- Argoverse2 · 2 of 17 papers
- SemanticKITTI · 2 of 17 papers
- nuScenes-C · 2 of 17 papers
- Waymo · 2 of 17 papers
Baselines
- TPVFormer · 3 of 17 papers
- BEVFormer · 3 of 17 papers
- No Collaboration · 2 of 17 papers
- Late Fusion · 2 of 17 papers
- V2VNet · 2 of 17 papers
- V2X-ViT · 2 of 17 papers
- CoAlign · 2 of 17 papers
- CoBEVT · 2 of 17 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 58 papers; from OpenAlex and the papers' affiliations).
- Huawei · 4 papers
- Tsinghua University · 4 papers
- The Chinese University of Hong Kong · 4 papers
- The University of Hong Kong · 3 papers
- University of Macau · 3 papers
- Beijing Institute of Technology · 3 papers
- Chinese Academy of Sciences · 3 papers
- Shanghai AI Laboratory · 3 papers
- University of Science and Technology of China · 2 papers
- Megvii · 2 papers
Use it
Download BibTeX of 35 key papersOpen the topic in the interactive map
Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.
Related approaches
Same keyword cluster
- Multi-camera BEV detection (69 papers)
- Camera-radar fusion detection (36 papers)
Elsewhere in the topic
- Unified LiDAR 3D detection (107 papers)
- Depth-based and label-free 3D detection (58 papers)
- 3D objectness and scene segmentation (57 papers)
- Monocular 3D object detection (56 papers)
Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).