mlsanityData to 14 Oct 2026

Hyperbolic Active Learning for Semantic Segmentation under Domain Shift

arXiv 2306.11180 ↗First posted 19 Jun 2023cs.CV cs.AIcited by 18

Not on the map. Why?

Abstract (arXiv)

We introduce a hyperbolic neural network approach to pixel-level active learning for semantic segmentation. Analysis of the data statistics leads to a novel interpretation of the hyperbolic radius as an indicator of data scarcity. In HALO (Hyperbolic Active Learning Optimization), for the first time, we propose the use of epistemic uncertainty as a data acquisition strategy, following the intuition of selecting data points that are the least known. The hyperbolic radius, complemented by the widely-adopted prediction entropy, effectively approximates epistemic uncertainty. We perform extensive experimental analysis based on two established synthetic-to-real benchmarks, i.e. GTAV rightarrow Cityscapes and SYNTHIA rightarrow Cityscapes. Additionally, we test HALO on Cityscape rightarrow ACDC for domain adaptation under adverse weather conditions, and we benchmark both convolutional and attention-based backbones. HALO sets a new state-of-the-art in active learning for semantic segmentation under domain shift and it is the first active learning approach that surpasses the performance of supervised domain adaptation while using only a small portion of labels (i.e., 1%).

Key facts. Cited by 18 papers in the map (data through 14 Oct 2026). Builds on 62 of its 83 references that are in the map. First posted on arXiv 19 Jun 2023.

Builds on

62 of 83 references.

References · 62

  1. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
  2. 2016The Cityscapes Dataset for Semantic Urban Scene Understandingcited by 4,444
  3. 2016Simple and Scalable Predictive Uncertainty Estimation using Deep Ensemblescited by 3,926
  4. 2016DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFscited by 3,241
  5. 2018Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentationcited by 3,223
  6. 2021SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformerscited by 2,940
  7. 2017What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?cited by 2,352
  8. 2017CyCADA: Cycle-Consistent Adversarial Domain Adaptationcited by 1,082
  9. 2019Aleatoric and Epistemic Uncertainty in Machine Learning: An Introduction to Concepts and Methodscited by 913
  10. 2017Deep Bayesian Active Learning with Image Datacited by 819
  11. 2016Playing for Data: Ground Truth from Computer Gamescited by 818
  12. 2017Poincar'e Embeddings for Learning Hierarchical Representationscited by 684
38 more
  1. 2018ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentationcited by 623
  2. 2019Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Boundscited by 610
  3. 2020FDA: Fourier Domain Adaptation for Semantic Segmentationcited by 541
  4. 2019BatchBALD: Efficient and Diverse Batch Acquisition for Deep Bayesian Active Learningcited by 432
  5. 2019Confidence Regularized Self-Trainingcited by 419
  6. 2018Hyperbolic Neural Networkscited by 402
  7. 2019Hyperbolic Graph Convolutional Neural Networkscited by 402
  8. 2019Semi-supervised Domain Adaptation via Minimax Entropycited by 335
  9. 2019Variational Adversarial Active Learningcited by 284
  10. 2021Prototypical Pseudo Label Denoising and Target Structure Learning for Domain Adaptive Semantic Segmentationcited by 268
  11. 2019Hyperbolic Graph Neural Networkscited by 244
  12. 2020Rectifying Pseudo Label Learning via Uncertainty Estimation for Domain Adaptive Semantic Segmentationcited by 222
  13. 2018Representation Tradeoffs for Hyperbolic Embeddingscited by 206
  14. 2022MIC: Masked Image Consistency for Context-Enhanced Domain Adaptationcited by 178
  15. 2020Instance Adaptive Self-Training for Unsupervised Domain Adaptationcited by 177
  16. 2021Source-Free Domain Adaptation for Semantic Segmentationcited by 168
  17. 2017Cost-Effective Active Learning for Deep Image Classificationcited by 166
  18. 2018Poincar'e GloVe: Hyperbolic Word Embeddingscited by 156
  19. 2018Hyperbolic Attention Networkscited by 122
  20. 2022Hyperbolic Vision Transformers: Combining Improvements in Metric Learningcited by 111
  21. 2022Hyperbolic Image Segmentationcited by 102
  22. 2022A Deeper Look into Aleatoric and Epistemic Uncertainty Disentanglementcited by 87
  23. 2021Generalize then Adapt: Source-Free Domain Adaptive Semantic Segmentationcited by 81
  24. 2018Quantifying Uncertainties in Natural Language Processing Taskscited by 79
  25. 2019Open Compound Domain Adaptationcited by 76
  26. 2019Active Adversarial Domain Adaptationcited by 73
  27. 2020Active Domain Adaptation via Clustering Uncertainty-weighted Embeddingscited by 72
  28. 2020From Trees to Continuous Embeddings and Back: Hyperbolic Hierarchical Clusteringcited by 65
  29. 2019Uncertainty Quantification with Statistical Guarantees in End-to-End Autonomous Driving Controlcited by 61
  30. 2020Two-phase Pseudo Label Densification for Self-training based Domain Adaptationcited by 60
  31. 2021Clipped Hyperbolic Classifiers Are Super-Hyperbolic Classifierscited by 57
  32. 2021Towards Fewer Annotations: Active Learning via Region Impurity and Prediction Uncertainty for Domain Adaptive Semantic Segmentationcited by 57
  33. 2020Domain Adaptive Semantic Segmentation Using Weak Labelscited by 56
  34. 2020Differentiating through the Fr'echet Meancited by 53
  35. 2021Active Learning for Domain Adaptation: An Energy-Based Approachcited by 53
  36. 2021CLDA: Contrastive Learning for Semi-Supervised Domain Adaptationcited by 52
  37. 2021ReDAL: Region-based and Diversity-aware Active Learning for Point Cloud Semantic Segmentationcited by 52
  38. 2021Learning the Predictability of the Futurecited by 49

50 most cited of 62 All 62 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 8; 2025: 3; 2026: 7.

18 papers.

Sort:

By area:

  1. 2024Compositional Entailment Learning for Hyperbolic Vision-Language Modelscited by 74
  2. 2024Hyperbolic Learning with Multimodal Large Language Modelscited by 10
  3. 2024Flattening the Parent Bias: Hierarchical Semantic Segmentation in the Poincar'e Ballcited by 9
  4. 2024Taxonomy-Aware Continual Semantic Segmentation in Hyperbolic Spaces for Open-World Perceptioncited by 8
  5. 2024Integrating Deep Metric Learning with Coreset for Active Learning in 3D Segmentationcited by 8
  6. 2025Curvature Learning for Generalization of Hyperbolic Neural Networkscited by 4
  7. 2024Adversarial Attacks on Hyperbolic Networkscited by 3
  8. 2025Modality Alignment across Trees on Heterogeneous Hyperbolic Manifoldscited by 3
  9. 2024SS-ADA: A Semi-Supervised Active Domain Adaptation Framework for Semantic Segmentationcited by 2
  10. 2026Uncertainty-guided Compositional Alignment with Part-to-Whole Semantic Representativeness in Hyperbolic Vision-Language Modelscited by 2
  11. 2024Hyp2Nav: Hyperbolic Planning and Curiosity for Crowd Navigationcited by 1
  12. 2025RAFT -- A Domain Adaptation Framework for RGB & LiDAR Semantic Segmentationcited by 0
6 more
  1. 2026Learning Action Hierarchies via Hybrid Geometric Diffusioncited by 0
  2. 2026Lorentz Framework for Semantic Segmentationcited by 0
  3. 2026Learning Taxonomic Trees with Hierarchical Representation Regularization for Large Multimodal Modelscited by 0
  4. 2026Hyperbolic Hierarchical Clustering for Visual Representation Learningcited by 0
  5. 2026ICM: Intra-class Mixing for Domain Adaptation in Adverse Weathercited by 0
  6. 2026TC-ADA: One-Shot Active Domain Adaptation for Semantic Segmentationcited by 0