mlsanityData to 7 Oct 2026

Spherical Transformer for LiDAR-based 3D Recognition

arXiv 2303.12766 ↗First posted 22 Mar 2023cs.CV cs.AIcited by 124

Not on the map. Why?

Abstract (arXiv)

LiDAR-based 3D point cloud recognition has benefited various applications. Without specially considering the LiDAR point distribution, most current methods suffer from information disconnection and limited receptive field, especially for the sparse distant points. In this work, we study the varying-sparsity distribution of LiDAR points and present SphereFormer to directly aggregate information from dense close points to the sparse distant ones. We design radial window self-attention that partitions the space into multiple non-overlapping narrow and long windows. It overcomes the disconnection issue and enlarges the receptive field smoothly and dramatically, which significantly boosts the performance of sparse distant points. Moreover, to fit the narrow and long windows, we propose exponential splitting to yield fine-grained position encoding and dynamic feature selection to increase model representation ability. Notably, our method ranks 1st on both nuScenes and SemanticKITTI semantic segmentation benchmarks with 81.9% and 74.8% mIoU, respectively. Also, we achieve the 3rd place on nuScenes object detection benchmark with 72.8% NDS and 68.5% mAP. Code is available at https://github.com/dvlab-research/SphereFormer.git.

Key facts. Cited by 124 papers in the map (data through 7 Oct 2026). Builds on 78 of its 90 references that are in the map. First posted on arXiv 22 Mar 2023.

Builds on

78 of 90 references.

References · 78

  1. 2017Attention Is All You Needcited by 59,932
  2. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
  3. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
  4. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,082
  5. 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,617
  6. 2020End-to-End Object Detection with Transformerscited by 7,384
  7. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,223
  8. 2020Training data-efficient image transformers & distillation through attentioncited by 4,560
  9. 2019nuScenes: A multimodal dataset for autonomous drivingcited by 4,485
  10. 2017PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Spacecited by 4,384
  11. 2020Deformable DETR: Deformable Transformers for End-to-End Object Detectioncited by 3,410
  12. 2018Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentationcited by 3,210
38 more
  1. 2016Pyramid Scene Parsing Networkcited by 2,472
  2. 2019Scalability in Perception for Autonomous Driving: Waymo Open Datasetcited by 2,436
  3. 2021Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutionscited by 1,960
  4. 2018PointPillars: Fast Encoders for Object Detection from Point Cloudscited by 1,707
  5. 2020Point Transformercited by 1,462
  6. 2017VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detectioncited by 1,323
  7. 2019KPConv: Flexible and Deformable Convolution for Point Cloudscited by 1,202
  8. 20194D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networkscited by 1,201
  9. 2020Center-based 3D Object Detection and Trackingcited by 1,143
  10. 2018PointRCNN: 3D Object Proposal Generation and Detection from Point Cloudcited by 939
  11. 2019PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detectioncited by 890
  12. 2021Going deeper with Image Transformerscited by 759
  13. 2021PVT v2: Improved Baselines with Pyramid Vision Transformercited by 749
  14. 20173D Semantic Segmentation with Submanifold Sparse Convolutional Networkscited by 746
  15. 2019RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Cloudscited by 584
  16. 2021Twins: Revisiting the Design of Spatial Attention in Vision Transformerscited by 565
  17. 2017Understanding the Effective Receptive Field in Deep Convolutional Neural Networkscited by 549
  18. 2021CSWin Transformer: A General Vision Transformer Backbone with Cross-Shaped Windowscited by 526
  19. 2022TransFusion: Robust LiDAR-Camera Fusion for 3D Object Detection with Transformerscited by 464
  20. 2018Spherical CNNscited by 452
  21. 20203DSSD: Point-based 3D Single Stage Object Detectorcited by 440
  22. 2020Searching Efficient 3D Architectures with Sparse Point-Voxel Convolutioncited by 422
  23. 2020Voxel R-CNN: Towards High Performance Voxel-based 3D Object Detectioncited by 409
  24. 2020Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR Segmentationcited by 388
  25. 2019Class-balanced Grouping and Sampling for Point Cloud 3D Object Detectioncited by 350
  26. 2020Exploring Self-attention for Image Recognitioncited by 313
  27. 2021Focal Self-attention for Local-Global Interactions in Vision Transformerscited by 286
  28. 2020PolarNet: An Improved Grid Representation for Online LiDAR Point Clouds Semantic Segmentationcited by 274
  29. 2021Voxel Transformer for 3D Object Detectioncited by 247
  30. 2020PointASNL: Robust Point Clouds Processing using Nonlocal Neural Networks with Adaptive Samplingcited by 238
  31. 2018SqueezeSegV2: Improved Model Structure and Unsupervised Domain Adaptation for Road-Object Segmentation from a LiDAR Point Cloudcited by 228
  32. 2020SalsaNext: Fast, Uncertainty-aware Semantic Segmentation of LiDAR Point Clouds for Autonomous Drivingcited by 226
  33. 2017Submanifold Sparse Convolutional Networkscited by 223
  34. 2020Sparse Single Sweep LiDAR Point Cloud Segmentation via Learning Contextual Shape Priors from Scene Completioncited by 222
  35. 2021Embracing Single Stride 3D Object Detector with Sparse Transformercited by 193
  36. 2018Tangent Convolutions for Dense Prediction in 3Dcited by 173
  37. 2021RPVNet: A Deep and Efficient Range-Point-Voxel Fusion Network for LiDAR Point Cloud Segmentationcited by 173
  38. 2020SqueezeSegV3: Spatially-Adaptive Convolution for Efficient Point-Cloud Segmentationcited by 170

50 most cited of 78 All 78 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 17; 2024: 41; 2025: 43; 2026: 23.

124 papers.

Sort:

By area:

  1. 2023Point Transformer V3: Simpler, Faster, Strongercited by 579
  2. 2023Towards Large-scale 3D Representation Learning with Multi-dataset Point Prompt Trainingcited by 85
  3. 2023UniPAD: A Universal Pre-training Paradigm for Autonomous Drivingcited by 73
  4. 2023PonderV2: Pave the Way for 3D Foundation Model with A Universal Pre-training Paradigmcited by 56
  5. 2023PointOcc: Cylindrical Tri-Perspective View for Point-based 3D Semantic Occupancy Predictioncited by 51
  6. 2023FRNet: Frustum-Range Networks for Scalable LiDAR Segmentationcited by 50
  7. 2023Mask-Attention-Free Transformer for 3D Instance Segmentationcited by 45
  8. 2023Three Pillars improving Vision Foundation Model Distillation for Lidarcited by 43
  9. 2024OA-CNNs: Omni-Adaptive Sparse CNNs for 3D Semantic Segmentationcited by 43
  10. 2023IST-Net: Prior-free Category-level Pose Estimation with Implicit Space Transformationcited by 40
  11. 2025DINO in the Room: Leveraging 2D Foundation Models for 3D Segmentationcited by 31
  12. 2024Multi-Space Alignments Towards Universal LiDAR Segmentationcited by 24
38 more
  1. 2024RadarPillars: Efficient Object Detection from 4D Radar Point Cloudscited by 23
  2. 2024An End-to-End Robust Point Cloud Semantic Segmentation Network with Single-Step Conditional Diffusion Modelscited by 22
  3. 2023HEAL-SWIN: A Vision Transformer On The Spherecited by 18
  4. 2025LitePT: Lighter Yet Stronger Point Transformercited by 17
  5. 2023ConDaFormer: Disassembled Transformer with Local Structure Enhancement for 3D Point Cloud Understandingcited by 16
  6. 2023WildScenes: A Benchmark for 2D and 3D Semantic Segmentation in Large-scale Natural Environmentscited by 16
  7. 2024MCD: Diverse Large-Scale Multi-Campus Dataset for Robot Perceptioncited by 15
  8. 2024LSK3DNet: Towards Effective and Efficient 3D Perception with Large Sparse Kernelscited by 15
  9. 2025Position: Beyond Euclidean -- Foundation Models Should Embrace Non-Euclidean Geometriescited by 15
  10. 2024Rethinking Data Augmentation for Robust LiDAR Semantic Segmentation in Adverse Weathercited by 14
  11. 2024TASeg: Temporal Aggregation Network for LiDAR Semantic Segmentationcited by 13
  12. 2024SFPNet: Sparse Focal Point Network for Semantic Segmentation on General LiDAR Point Cloudscited by 13
  13. 2024Efficient 3D Recognition with Event-driven Spike Sparse Convolutioncited by 11
  14. 2024A Point-Based Approach to Efficient LiDAR Multi-Task Perceptioncited by 9
  15. 2024Building a Strong Pre-Training Baseline for Universal 3D Large-Scale Perceptioncited by 8
  16. 2025LeAP: Consistent multi-domain 3D labeling using Foundation Modelscited by 8
  17. 2025Label-Efficient LiDAR Semantic Segmentation with 2D-3D Vision Transformer Adapterscited by 8
  18. 2023PeP: a Point enhanced Painting method for unified point cloud taskscited by 7
  19. 2024Density-guided Translator Boosts Synthetic-to-Real Unsupervised Domain Adaptive Segmentation of 3D Point Cloudscited by 7
  20. 2024SALSA: Swift Adaptive Lightweight Self-Attention for Enhanced LiDAR Place Recognitioncited by 7
  21. 2024TripleMixer: A 3D Point Cloud Denoising Model for Adverse Weathercited by 7
  22. 2025Multi-Scale Neighborhood Occupancy Masked Autoencoder for Self-Supervised Learning in LiDAR Point Cloudscited by 7
  23. 2023Spherical Frustum Sparse Convolution Network for LiDAR Point Cloud Semantic Segmentationcited by 5
  24. 2024SegNet4D: Efficient Instance-Aware 4D Semantic Segmentation for LiDAR Point Cloudcited by 5
  25. 2024ItTakesTwo: Leveraging Peer Representations for Semi-supervised LiDAR Semantic Segmentationcited by 5
  26. 2025No Thing, Nothing: Highlighting Safety-Critical Classes for Robust LiDAR Semantic Segmentation in Adverse Weathercited by 5
  27. 2025NUC-Net: Non-uniform Cylindrical Partition Network for Efficient LiDAR Semantic Segmentationcited by 5
  28. 2024PoIFusion: Multi-Modal 3D Object Detection via Fusion at Points of Interestcited by 4
  29. 2024PDF: A Probability-Driven Framework for Open World 3D Point Cloud Semantic Segmentationcited by 4
  30. 2024Filling Missing Values Matters for Range Image-Based Point Cloud Segmentationcited by 4
  31. 2024Serialized Point Mamba: A Serialized Point Cloud Mamba Segmentation Modelcited by 4
  32. 2024Panoptic Perception for Autonomous Driving: A Surveycited by 4
  33. 2024Exploiting Local Features and Range Images for Small Data Real-Time Point Cloud Semantic Segmentationcited by 4
  34. 2025SALT: A Flexible Semi-Automatic Labeling Tool for General LiDAR Point Clouds with Cross-Scene Adaptability and 4D Consistencycited by 4
  35. 2025How Many Tokens Do 3D Point Cloud Transformer Architectures Really Need?cited by 4
  36. 2026Volume Transformer: Revisiting Vanilla Transformers for 3D Scene Understandingcited by 4
  37. 2023BEVSeg2TP: Surround View Camera Bird's-Eye-View Based Joint Vehicle Segmentation and Ego Vehicle Trajectory Predictioncited by 3
  38. 2024AutoInst: Automatic Instance-Based Segmentation of LiDAR 3D Scanscited by 3

50 most cited of 124 All 124 (JSON)