mlsanityData to 7 Oct 2026

Infinite Mixture Prototypes for Few-Shot Learning

arXiv 1902.04552 ↗First posted 12 Feb 2019cs.LG stat.MLcited by 99

Not on the map. Why?

Abstract (arXiv)

We propose infinite mixture prototypes to adaptively represent both simple and complex data distributions for few-shot learning. Our infinite mixture prototypes represent each class by a set of clusters, unlike existing prototypical methods that represent each class by a single cluster. By inferring the number of clusters, infinite mixture prototypes interpolate between nearest neighbor and prototypical representations, which improves accuracy and robustness in the few-shot regime. We show the importance of adaptive capacity for capturing complex data distributions such as alphabets, with 25% absolute accuracy improvements over prototypical networks, while still maintaining or improving accuracy on the standard Omniglot and mini-ImageNet benchmarks. In clustering labeled and unlabeled data by the same clustering rule, infinite mixture prototypes achieves state-of-the-art semi-supervised accuracy. As a further capability, we show that infinite mixture prototypes can perform purely unsupervised clustering, unlike existing prototypical methods.

Key facts. Cited by 99 papers in the map (data through 7 Oct 2026). First posted on arXiv 12 Feb 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 35; 2022: 22; 2023: 16; 2024: 11; 2025: 13; 2026: 2.

99 papers.

Sort:

By area:

  1. 2021Hypercorrelation Squeeze for Few-Shot Segmentationcited by 193
  2. 2022Self-Support Few-Shot Semantic Segmentationcited by 96
  3. 2021Relational Embedding for Few-Shot Classificationcited by 74
  4. 2023Advances and Challenges in Meta-Learning: A Technical Reviewcited by 65
  5. 2021Exploring Complementary Strengths of Invariant and Equivariant Representations for Few-Shot Learningcited by 56
  6. 2022HyperTransformer: Model Generation for Supervised and Semi-Supervised Few-Shot Learningcited by 55
  7. 2021von Mises-Fisher Loss: An Exploration of Embedding Geometries for Supervised Learningcited by 42
  8. 2023A General-Purpose Self-Supervised Model for Computational Pathologycited by 42
  9. 2022Novel Class Discovery without Forgettingcited by 39
  10. 2021Exemplar-Based Open-Set Panoptic Segmentation Networkcited by 37
  11. 2023On the Robustness of Open-World Test-Time Training: Self-Training with Dynamic Prototype Expansioncited by 34
  12. 2021A Metalearning Approach for Physics-Informed Neural Networks (PINNs): Application to Parameterized PDEscited by 31
38 more
  1. 2022Integrative Few-Shot Learning for Classification and Segmentationcited by 25
  2. 2022Matching Feature Sets for Few-Shot Image Classificationcited by 25
  3. 2021Deep Metric Learning for Few-Shot Image Classification: A Review of Recent Developmentscited by 23
  4. 2021Prototype Guided Federated Learning of Visual Feature Representationscited by 22
  5. 2024Prototypical Reward Network for Data-Efficient RLHFcited by 22
  6. 2023Understanding the (Extra-)Ordinary: Validating Deep Model Decisions with Prototypical Concept-based Explanationscited by 21
  7. 2023Context-Aware Meta-Learningcited by 20
  8. 2025Model Predictive Task Sampling for Efficient and Robust Adaptationcited by 20
  9. 2021Learning Prototype-oriented Set Representations for Meta-Learningcited by 19
  10. 2023MaskDiff: Modeling Mask Distribution with Diffusion Probabilistic Model for Few-Shot Instance Segmentationcited by 18
  11. 2021Dense Gaussian Processes for Few-Shot Segmentationcited by 17
  12. 2021Temporal Action Segmentation with High-level Complex Activity Labelscited by 16
  13. 2021Distance-wise Prototypical Graph Neural Network in Node Imbalance Classificationcited by 16
  14. 2021Long-tail Recognition via Compositional Knowledge Transfercited by 16
  15. 2021Contrastive Prototype Learning with Augmented Embeddings for Few-Shot Learningcited by 15
  16. 2021A Closer Look at Prototype Classifier for Few-shot Image Classificationcited by 15
  17. 2021Hybrid Graph Neural Networks for Few-Shot Learningcited by 15
  18. 2023ProtoDiff: Learning to Learn Prototypical Networks by Task-Guided Diffusioncited by 15
  19. 2021Hierarchical Variational Memory for Few-shot Learning Across Domainscited by 14
  20. 2022Weakly Supervised 3D Point Cloud Segmentation via Multi-Prototype Learningcited by 13
  21. 2024Robust Fast Adaptation from Adversarially Explicit Task Distribution Generationcited by 12
  22. 2021Multi-Label Few-Shot Learning for Aspect Category Detectioncited by 11
  23. 2023Hubs and Hyperspheres: Reducing Hubness and Improving Transductive Few-shot Learning with Hyperspherical Embeddingscited by 11
  24. 2022Prototypical Fine-tuning: Towards Robust Performance Under Varying Data Sizescited by 10
  25. 2023EEGMatch: Learning with Incomplete Labels for Semi-Supervised EEG-based Cross-Subject Emotion Recognitioncited by 10
  26. 2022Few-Shot Document-Level Event Argument Extractioncited by 9
  27. 2023A Simple Yet Effective Strategy to Robustify the Meta Learning Paradigmcited by 9
  28. 2021Bayesian Embeddings for Few-Shot Open World Recognitioncited by 8
  29. 2022Federated Semi-Supervised Learning with Prototypical Networkscited by 8
  30. 2023Episodic Multi-Task Learning with Heterogeneous Neural Processescited by 8
  31. 2021On the Importance of Firth Bias Reduction in Few-Shot Classificationcited by 7
  32. 2022Label-enhanced Prototypical Network with Contrastive Learning for Multi-label Few-shot Aspect Category Detectioncited by 7
  33. 2023In-Context Learning for Few-Shot Molecular Property Predictioncited by 7
  34. 2025Uncertainty Meets Diversity: A Comprehensive Active Learning Framework for Indoor 3D Object Detectioncited by 7
  35. 2022Meta Learning of Interface Conditions for Multi-Domain Physics-Informed Neural Networkscited by 6
  36. 2024Localization and Expansion: A Decoupled Framework for Point Cloud Few-shot Semantic Segmentationcited by 6
  37. 2021Omni-Training: Bridging Pre-Training and Meta-Training for Few-Shot Learningcited by 5
  38. 2021Meta-Learning with Adjoint Methodscited by 5

50 most cited of 99 All 99 (JSON)