mlsanityData to 7 Oct 2026

Voxel R-CNN: Towards High Performance Voxel-based 3D Object Detection

arXiv 2012.15712 ↗First posted 31 Dec 2020cs.CVcited by 409

Not on the map. Why?

Abstract (arXiv)

Recent advances on 3D object detection heavily rely on how the 3D data are represented, i.e., voxel-based or point-based representation. Many existing high performance 3D detectors are point-based because this structure can better retain precise point positions. Nevertheless, point-level features lead to high computation overheads due to unordered storage. In contrast, the voxel-based structure is better suited for feature extraction but often yields lower accuracy because the input data are divided into grids. In this paper, we take a slightly different viewpoint -- we find that precise positioning of raw points is not essential for high performance 3D object detection and that the coarse voxel granularity can also offer sufficient detection accuracy. Bearing this view in mind, we devise a simple but effective voxel-based framework, named Voxel R-CNN. By taking full advantage of voxel features in a two stage approach, our method achieves comparable detection accuracy with state-of-the-art point-based models, but at a fraction of the computation cost. Voxel R-CNN consists of a 3D backbone network, a 2D bird-eye-view (BEV) Region Proposal Network and a detect head. A voxel RoI pooling is devised to extract RoI features directly from voxel features for further refinement. Extensive experiments are conducted on the widely used KITTI Dataset and the more recent Waymo Open Dataset. Our results show that compared to existing voxel-based methods, Voxel R-CNN delivers a higher detection accuracy while maintaining a real-time frame processing rate, i.e., at a speed of 25 FPS on an NVIDIA RTX 2080 Ti GPU. The code is available at urlhttps://github.com/djiajunustc/Voxel-R-CNN.

Key facts. Cited by 409 papers in the map (data through 7 Oct 2026). First posted on arXiv 31 Dec 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 26; 2022: 89; 2023: 110; 2024: 91; 2025: 64; 2026: 29.

409 papers.

Sort:

By area:

  1. 2021Voxel Transformer for 3D Object Detectioncited by 247
  2. 2021PV-RCNN++: Point-Voxel Feature Set Abstraction With Local Vector Representation for 3D Object Detectioncited by 232
  3. 2023VoxelNeXt: Fully Sparse VoxelNet for 3D Object Detection and Trackingcited by 221
  4. 2022Unifying Voxel-based Representation with Transformer for 3D Object Detectioncited by 202
  5. 2021Embracing Single Stride 3D Object Detector with Sparse Transformercited by 193
  6. 2023Think Twice before Driving: Towards Scalable Decoders for End-to-End Autonomous Drivingcited by 168
  7. 2022Focal Sparse Convolutional Networks for 3D Object Detectioncited by 142
  8. 2021RangeDet:In Defense of Range View for LiDAR-based 3D Object Detectioncited by 140
  9. 2024ShapeLLM: Universal 3D Object Understanding for Embodied Interactioncited by 133
  10. 20223D Object Detection for Autonomous Driving: A Comprehensive Surveycited by 132
  11. 2022Not All Points Are Equal: Learning Highly Efficient Point-based Detectors for 3D LiDAR Point Cloudscited by 131
  12. 2023Robo3D: Towards Robust and Reliable 3D Perception against Corruptionscited by 128
38 more
  1. 2023Spherical Transformer for LiDAR-based 3D Recognitioncited by 124
  2. 2022Delving into the Devils of Bird's-eye-view Perception: A Review, Evaluation and Recipecited by 122
  3. 2023DSVT: Dynamic Sparse Voxel Transformer with Rotated Setscited by 122
  4. 2021Improving 3D Object Detection with Channel-wise Transformercited by 108
  5. 20213D Object Detection for Autonomous Driving: A Surveycited by 106
  6. 2022Voxel Set Transformer: A Set-to-Set Approach to 3D Object Detection from Point Cloudscited by 102
  7. 2022PillarNet: Real-Time and High-Performance Pillar-based 3D Object Detectioncited by 99
  8. 2022CenterFormer: Center-based Transformer for 3D Object Detectioncited by 99
  9. 2021SPG: Unsupervised Domain Adaptation for 3D Object Detection via Semantic Point Generationcited by 93
  10. 2021AFDetV2: Rethinking the Necessity of the Second Stage for Object Detection from Point Cloudscited by 88
  11. 2022Fully Sparse 3D Object Detectioncited by 84
  12. 2023SparseFusion: Fusing Multi-Modal Sparse Representations for Multi-Sensor 3D Object Detectioncited by 83
  13. 2022Sparse Fuse Dense: Towards High Quality 3D Detection with Depth Completioncited by 82
  14. 2023Virtual Sparse Convolution for Multimodal 3D Object Detectioncited by 81
  15. 2023Collaboration Helps Camera Overtake LiDAR in 3D Detectioncited by 81
  16. 2023LoGoNet: Towards Accurate 3D Object Detection with Local-to-Global Cross-Modal Fusioncited by 80
  17. 2021Pyramid R-CNN: Towards Better Performance and Adaptability for 3D Object Detectioncited by 79
  18. 2023UniTR: A Unified and Efficient Multi-Modal Transformer for Bird's-Eye-View Representationcited by 74
  19. 2022Multi-modal Sensor Fusion for Auto Driving Perception: A Surveycited by 73
  20. 2023UniPAD: A Universal Pre-training Paradigm for Autonomous Drivingcited by 73
  21. 2022LiDAR Snowfall Simulation for Robust 3D Object Detectioncited by 72
  22. 2022LargeKernel3D: Scaling up Kernels in 3D Sparse CNNscited by 71
  23. 2022Point Density-Aware Voxels for LiDAR 3D Object Detectioncited by 70
  24. 2024Voxel Mamba: Group-Free State Space Models for Point Cloud based 3D Object Detectioncited by 66
  25. 2021Multi-Modal 3D Object Detection in Autonomous Driving: a Surveycited by 63
  26. 2021Behind the Curtain: Learning Occluded Shapes for 3D Object Detectioncited by 58
  27. 2022CAGroup3D: Class-Aware Grouping for 3D Object Detection on Point Cloudscited by 58
  28. 2023FocalFormer3D : Focusing on Hard Instance for 3D Object Detectioncited by 58
  29. 2022GD-MAE: Generative Decoder for MAE Pre-training on LiDAR Point Cloudscited by 57
  30. 2022LiDAR Distillation: Bridging the Beam-Induced Domain Gap for 3D Object Detectioncited by 54
  31. 2022SASA: Semantics-Augmented Set Abstraction for Point-based 3D Object Detectioncited by 52
  32. 2022Transformation-Equivariant 3D Object Detection for Autonomous Drivingcited by 52
  33. 2022Voxel Field Fusion for 3D Object Detectioncited by 50
  34. 2021EPNet++: Cascade Bi-directional Fusion for Multi-Modal 3D Object Detectioncited by 49
  35. 2021VPFNet: Improving 3D Object Detection with Virtual Point based LiDAR and Stereo Data Fusioncited by 48
  36. 2022Cross-Modality Knowledge Distillation Network for Monocular 3D Object Detectioncited by 48
  37. 2022AutoAlignV2: Deformable Feature Aggregation for Dynamic Multi-Modal 3D Object Detectioncited by 46
  38. 2025SoFar: Language-Grounded Orientation Bridges Spatial Reasoning and Object Manipulationcited by 46

50 most cited of 409 All 409 (JSON)