mlsanityData to 14 Oct 2026

You Never Get a Second Chance To Make a Good First Impression: Seeding Active Learning for 3D Semantic Segmentation

arXiv 2304.11762 ↗First posted 23 Apr 2023cs.CVcited by 6

Not on the map. Why?

Abstract (arXiv)

We propose SeedAL, a method to seed active learning for efficient annotation of 3D point clouds for semantic segmentation. Active Learning (AL) iteratively selects relevant data fractions to annotate within a given budget, but requires a first fraction of the dataset (a 'seed') to be already annotated to estimate the benefit of annotating other data fractions. We first show that the choice of the seed can significantly affect the performance of many AL methods. We then propose a method for automatically constructing a seed that will ensure good performance for AL. Assuming that images of the point clouds are available, which is common, our method relies on powerful unsupervised image features to measure the diversity of the point clouds. It selects the point clouds for the seed by optimizing the diversity under an annotation budget, which can be done by solving a linear optimization problem. Our experiments demonstrate the effectiveness of our approach compared to random seeding and existing methods on both the S3DIS and SemanticKitti datasets. Code is available at https://github.com/nerminsamet/seedal.

Key facts. Cited by 6 papers in the map (data through 14 Oct 2026). Builds on 64 of its 87 references that are in the map. First posted on arXiv 23 Apr 2023.

Builds on

64 of 87 references.

References · 64

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,399
  2. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  3. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
  4. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,513
  5. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,968
  6. 2020Bootstrap your own latent: A new approach to self-supervised Learningcited by 4,750
  7. 2019nuScenes: A multimodal dataset for autonomous drivingcited by 4,523
  8. 2017ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenescited by 3,117
  9. 2020Improved Baselines with Momentum Contrastive Learningcited by 2,287
  10. 2021An Empirical Study of Training Self-Supervised Vision Transformerscited by 1,616
  11. 2020Point Transformercited by 1,470
  12. 2019KPConv: Flexible and Deformable Convolution for Point Cloudscited by 1,205
38 more
  1. 2017Deep Bayesian Active Learning with Image Datacited by 819
  2. 2019Self-supervised Visual Feature Learning with Deep Neural Networks: A Surveycited by 657
  3. 2019Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Boundscited by 610
  4. 2020A Survey of Deep Active Learningcited by 554
  5. 2020PointContrast: Unsupervised Pre-training for 3D Point Cloud Understandingcited by 504
  6. 2022PointNeXt: Revisiting PointNet++ with Improved Training and Scaling Strategiescited by 436
  7. 2020Searching Efficient 3D Architectures with Sparse Point-Voxel Convolutioncited by 423
  8. 2020Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR Segmentationcited by 389
  9. 2021Localizing Objects with Self-Supervised Transformers and no Labelscited by 230
  10. 2021Self-Supervised Pretraining of 3D Features on any Point-Cloudcited by 217
  11. 2021Spatio-temporal Self-Supervised Representation Learning for 3D Point Cloudscited by 180
  12. 2022Self-Supervised Transformers for Unsupervised Object Discovery using Normalized Cutcited by 167
  13. 2021ST3D: Self-training for Unsupervised Domain Adaptation on 3D Object Detectioncited by 159
  14. 20222DPASS: 2D Priors Assisted Semantic Segmentation on LiDAR Point Cloudscited by 158
  15. 2022Active Learning on a Budget: Opposite Strategies Suit High and Low Budgetscited by 133
  16. 2020Cold-start Active Learning through Self-supervised Language Modelingcited by 128
  17. 2022Image-to-Lidar Self-Supervised Distillation for Autonomous Driving Datacited by 126
  18. 2022Point-to-Voxel Knowledge Distillation for LiDAR Semantic Segmentationcited by 119
  19. 2020Complete & Label: A Domain Adaptation Approach to Semantic Segmentation of LiDAR Point Cloudscited by 116
  20. 2019Consistency-based Semi-supervised Active Learning: Towards Minimizing Labeling Costcited by 114
  21. 2017Learning Active Learning from Datacited by 101
  22. 2021SPG: Unsupervised Domain Adaptation for 3D Object Detection via Semantic Point Generationcited by 94
  23. 2019Deep Active Learning: Unified and Principled Method for Query and Trainingcited by 93
  24. 2021One Thing One Click: A Self-Training Approach for Weakly Supervised 3D Semantic Segmentationcited by 89
  25. 2022Learning from Few Examples: A Summary of Approaches to Few-Shot Learningcited by 86
  26. 2017Deep Active Learning over the Long Tailcited by 84
  27. 2020Multi-Path Region Mining For Weakly Supervised 3D Semantic Segmentation on Point Cloudscited by 71
  28. 2021PointMixer: MLP-Mixer for Point Cloud Understandingcited by 71
  29. 2022ALSO: Automotive Lidar Self-supervision by Occupancy estimationcited by 68
  30. 2022Active Learning Through a Covering Lenscited by 66
  31. 2020OBoW: Online Bag-of-Visual-Words Generation for Self-Supervised Learningcited by 65
  32. 2021Learning 3D Semantic Segmentation with only 2D Image Supervisioncited by 60
  33. 2021SSPC-Net: Semi-supervised Semantic 3D Point Cloud Segmentation Networkcited by 54
  34. 2022CoSMix: Compositional Semantic Mix for Domain Adaptation in 3D LiDAR Segmentationcited by 54
  35. 2021ReDAL: Region-based and Diversity-aware Active Learning for Point Cloud Semantic Segmentationcited by 52
  36. 2018Active Learning for Deep Object Detectioncited by 50
  37. 2019Deep Active Learning for Efficient Training of a LiDAR 3D Object Detectorcited by 45
  38. 2022LESS: Label-Efficient Semantic Segmentation for LiDAR Point Cloudscited by 41

50 most cited of 64 All 64 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 1; 2024: 2; 2025: 2; 2026: 1.

6 papers.

  1. 2025Perspective-Invariant 3D Object Detectioncited by 18
  2. 2023Annotator: A Generic Active Learning Baseline for LiDAR Semantic Segmentationcited by 11
  3. 2024LOGen: Toward Lidar Object Generation by Point Diffusioncited by 9
  4. 2024MILAN: Milli-Annotations for Lidar Semantic Segmentationcited by 0
  5. 2025Label-Efficient Point Cloud Segmentation with Active Learningcited by 0
  6. 2026Energy-Based Open-Set Active Learning for Object Classificationcited by 0