mlsanityData to 14 Oct 2026

LiDAL: Inter-frame Uncertainty Based Active Learning for 3D LiDAR Semantic Segmentation

arXiv 2211.05997 ↗First posted 11 Nov 2022cs.CVcited by 16

Not on the map. Why?

Abstract (arXiv)

We propose LiDAL, a novel active learning method for 3D LiDAR semantic segmentation by exploiting inter-frame uncertainty among LiDAR frames. Our core idea is that a well-trained model should generate robust results irrespective of viewpoints for scene scanning and thus the inconsistencies in model predictions across frames provide a very reliable measure of uncertainty for active sample selection. To implement this uncertainty measure, we introduce new inter-frame divergence and entropy formulations, which serve as the metrics for active selection. Moreover, we demonstrate additional performance gains by predicting and incorporating pseudo-labels, which are also selected using the proposed inter-frame uncertainty measure. Experimental results validate the effectiveness of LiDAL: we achieve 95% of the performance of fully supervised learning with less than 5% of annotations on the SemanticKITTI and nuScenes datasets, outperforming state-of-the-art active learning methods. Code release: https://github.com/hzykent/LiDAL.

Key facts. Cited by 16 papers in the map (data through 14 Oct 2026). Builds on 35 of its 60 references that are in the map. First posted on arXiv 11 Nov 2022.

Builds on

35 of 60 references.

References · 35

  1. 2019nuScenes: A multimodal dataset for autonomous drivingcited by 4,523
  2. 20194D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networkscited by 1,208
  3. 2019KPConv: Flexible and Deformable Convolution for Point Cloudscited by 1,205
  4. 2019RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Cloudscited by 585
  5. 2020PointContrast: Unsupervised Pre-training for 3D Point Cloud Understandingcited by 504
  6. 2020Searching Efficient 3D Architectures with Sparse Point-Voxel Convolutioncited by 423
  7. 2020Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR Segmentationcited by 389
  8. 2018SqueezeSegV2: Improved Model Structure and Unsupervised Domain Adaptation for Road-Object Segmentation from a LiDAR Point Cloudcited by 228
  9. 2021Self-Supervised Pretraining of 3D Features on any Point-Cloudcited by 217
  10. 2020Exploring Data-Efficient 3D Scene Understanding with Contrastive Scene Contextscited by 210
  11. 2020Unsupervised Point Cloud Pre-Training via Occlusion Completioncited by 186
  12. 2021Spatio-temporal Self-Supervised Representation Learning for 3D Point Cloudscited by 180
23 more
  1. 2017Cost-Effective Active Learning for Deep Image Classificationcited by 166
  2. 2020Complete & Label: A Domain Adaptation Approach to Semantic Segmentation of LiDAR Point Cloudscited by 116
  3. 2016Vehicle Detection from 3D Lidar Using Fully Convolutional Networkcited by 107
  4. 2019ViewAL: Active Learning with Viewpoint Entropy for Semantic Segmentationcited by 96
  5. 2019Unsupervised Multi-Task Feature Learning on Point Cloudscited by 89
  6. 2021One Thing One Click: A Self-Training Approach for Weakly Supervised 3D Semantic Segmentationcited by 89
  7. 2020Multi-Path Region Mining For Weakly Supervised 3D Semantic Segmentation on Point Cloudscited by 71
  8. 2020Self-Supervised Few-Shot Learning on Point Cloudscited by 66
  9. 2021SQN: Weakly-Supervised Semantic Segmentation of Large-Scale 3D Point Cloudscited by 64
  10. 2020JSENet: Joint Semantic Segmentation and Edge Detection Network for 3D Point Cloudscited by 62
  11. 2021SSPC-Net: Semi-supervised Semantic 3D Point Cloud Segmentation Networkcited by 54
  12. 2021ReDAL: Region-based and Diversity-aware Active Learning for Point Cloud Semantic Segmentationcited by 52
  13. 2018CEREALS - Cost-Effective REgion-based Active Learning for Semantic Segmentationcited by 49
  14. 2017Cost-Effective Active Learning for Melanoma Segmentationcited by 41
  15. 2020P4Contrast: Contrastive Learning with Pairs of Point-Pixel Pairs for RGB-D Scene Understandingcited by 41
  16. 20213D Spatial Recognition without Spatially Labeled 3Dcited by 31
  17. 2020Canonical Capsules: Self-Supervised Capsules in Canonical Posecited by 30
  18. 2020SegGroup: Seg-Level Supervision for 3D Instance and Semantic Segmentationcited by 30
  19. 2020Weakly Supervised Semantic Segmentation in 3D Graph-Structured Point Clouds of Wild Scenescited by 27
  20. 2021Label-Efficient Point Cloud Semantic Segmentation: An Active Learning Approachcited by 17
  21. 2022Box2Seg: Learning Semantics of 3D Point Clouds with Box-Level Supervisioncited by 11
  22. 2019MortonNet: Self-Supervised Learning of Local Features in 3D Point Cloudscited by 9
  23. 2021Dense Supervision Propagation for Weakly Supervised Semantic Segmentation on 3D Point Cloudscited by 4

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 10; 2024: 2; 2025: 2; 2026: 2.

16 papers.

Sort:

  1. 2023Segment Any Point Cloud Sequences by Distilling Vision Foundation Modelscited by 125
  2. 2023GrowSP: Unsupervised Semantic Segmentation of 3D Point Cloudscited by 34
  3. 2023Less is More: Reducing Task and Model Complexity for 3D Point Cloud Semantic Segmentationcited by 33
  4. 20244D Contrastive Superflows are Dense 3D Representation Learnerscited by 33
  5. 2023A Survey of Label-Efficient Deep Learning for 3D Point Cloudscited by 20
  6. 2023360^circ from a Single Camera: A Few-Shot Approach for LiDAR Segmentationcited by 12
  7. 2025LargeAD: Large-Scale Cross-Sensor Data Pretraining for Autonomous Drivingcited by 12
  8. 2023Annotator: A Generic Active Learning Baseline for LiDAR Semantic Segmentationcited by 11
  9. 2023BaSAL: Size-Balanced Warm Start Active Learning for LiDAR Semantic Segmentationcited by 8
  10. 2023Discwise Active Learning for LiDAR Semantic Segmentationcited by 7
  11. 2023You Never Get a Second Chance To Make a Good First Impression: Seeding Active Learning for 3D Semantic Segmentationcited by 6
  12. 2025LiDAR Remote Sensing Meets Weak Supervision: Concepts, Methods, and Perspectivescited by 4
4 more
  1. 2023iDet3D: Towards Efficient Interactive Object Detection for LiDAR Point Cloudscited by 3
  2. 2026Class-Distribution Guided Active Learning for 3D Occupancy Prediction in Autonomous Drivingcited by 1
  3. 2026Depth-Only Open-Vocabulary 3D Semantic Segmentation For Privacy-Preserving Robotic Applicationscited by 1
  4. 2024MILAN: Milli-Annotations for Lidar Semantic Segmentationcited by 0