mlsanityData to 14 Oct 2026

PVTransformer: Point-to-Voxel Transformer for Scalable 3D Object Detection

arXiv 2405.02811 ↗First posted 5 May 2024cs.CVcited by 3

Not on the map. Why?

Abstract (arXiv)

3D object detectors for point clouds often rely on a pooling-based PointNet to encode sparse points into grid-like voxels or pillars. In this paper, we identify that the common PointNet design introduces an information bottleneck that limits 3D object detection accuracy and scalability. To address this limitation, we propose PVTransformer: a transformer-based point-to-voxel architecture for 3D detection. Our key idea is to replace the PointNet pooling operation with an attention module, leading to a better point-to-voxel aggregation function. Our design respects the permutation invariance of sparse 3D points while being more expressive than the pooling-based PointNet. Experimental results show our PVTransformer achieves much better performance compared to the latest 3D object detectors. On the widely used Waymo Open Dataset, our PVTransformer achieves state-of-the-art 76.5 mAPH L2, outperforming the prior art of SWFormer by +1.7 mAPH L2.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 31 of its 34 references that are in the map. First posted on arXiv 5 May 2024.

Builds on

31 of 34 references.

References · 31

  1. 2017Attention Is All You Needcited by 60,361
  2. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,248
  3. 2019nuScenes: A multimodal dataset for autonomous drivingcited by 4,523
  4. 2017PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Spacecited by 4,401
  5. 2019Scalability in Perception for Autonomous Driving: Waymo Open Datasetcited by 2,451
  6. 2018PointPillars: Fast Encoders for Object Detection from Point Cloudscited by 1,715
  7. 2020Point Transformercited by 1,470
  8. 2017VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detectioncited by 1,323
  9. 2020Center-based 3D Object Detection and Trackingcited by 1,144
  10. 2021Perceiver: General Perception with Iterative Attentioncited by 1,104
  11. 2018PointRCNN: 3D Object Proposal Generation and Detection from Point Cloudcited by 941
  12. 2017Frustum PointNets for 3D Object Detection from RGB-D Datacited by 640
19 more
  1. 2019Point-Voxel CNN for Efficient 3D Deep Learningcited by 343
  2. 2021An End-to-End Transformer Model for 3D Object Detectioncited by 329
  3. 2021PV-RCNN++: Point-Voxel Feature Set Abstraction With Local Vector Representation for 3D Object Detectioncited by 232
  4. 2021Embracing Single Stride 3D Object Detector with Sparse Transformercited by 193
  5. 2019End-to-End Multi-View Fusion for 3D Object Detection in LiDAR Point Cloudscited by 171
  6. 2019Modeling Point Clouds with Self-Attention and Gumbel Subset Samplingcited by 151
  7. 2021RangeDet:In Defense of Range View for LiDAR-based 3D Object Detectioncited by 140
  8. 2019LaserNet: An Efficient Probabilistic 3D Object Detector for Autonomous Drivingcited by 133
  9. 2021LiDAR R-CNN: An Efficient and Universal 3D Object Detectorcited by 129
  10. 2021RSN: Range Sparse Net for Efficient, Accurate LiDAR 3D Object Detectioncited by 115
  11. 2018Point2Sequence: Learning the Shape Representation of 3D Point Clouds with an Attention-based Sequence to Sequence Networkcited by 100
  12. 2022CenterFormer: Center-based Transformer for 3D Object Detectioncited by 99
  13. 2022SWFormer: Sparse Window Transformer for 3D Object Detection in Point Cloudscited by 92
  14. 2021AFDetV2: Rethinking the Necessity of the Second Stage for Object Detection from Point Cloudscited by 88
  15. 2020Range Conditioned Dilated Convolutions for Scale Invariant 3D Object Detectioncited by 78
  16. 2019StarNet: Targeted Computation for Object Detection in Point Cloudscited by 46
  17. 2021To the Point: Efficient 3D Object Detection in the Range Image with Graph Convolution Kernelscited by 39
  18. 2020PV-RCNN: The Top-Performing LiDAR-only Solutions for 3D Detection / 3D Tracking / Domain Adaptation of Waymo Open Dataset Challengescited by 8
  19. 2022LidarAugment: Searching for Scalable 3D LiDAR Data Augmentationscited by 8

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 1; 2025: 1; 2026: 1.

3 papers.

  1. 2025DiffPixelFormer: Differential Pixel-Aware Transformer for RGB-D Indoor Scene Segmentationcited by 2
  2. 2024Managing Bandwidth: The Key to Cloud-Assisted Autonomous Drivingcited by 1
  3. 2026DRIFT: Dual-Representation Inter-Fusion Transformer for Automated Driving Perception with 4D Radar Point Cloudscited by 0