mlsanityData to 14 Oct 2026

The Role of Global Labels in Few-Shot Classification and How to Infer Them

arXiv 2108.04055 ↗First posted 9 Aug 2021cs.LG stat.MLcited by 12

Not on the map. Why?

Abstract (arXiv)

Few-shot learning is a central problem in meta-learning, where learners must quickly adapt to new tasks given limited training data. Recently, feature pre-training has become a ubiquitous component in state-of-the-art meta-learning methods and is shown to provide significant performance improvement. However, there is limited theoretical understanding of the connection between pre-training and meta-learning. Further, pre-training requires global labels shared across tasks, which may be unavailable in practice. In this paper, we show why exploiting pre-training is theoretically advantageous for meta-learning, and in particular the critical role of global labels. This motivates us to propose Meta Label Learning (MeLa), a novel meta-learning framework that automatically infers global labels to obtains robust few-shot models. Empirically, we demonstrate that MeLa is competitive with existing methods and provide extensive ablation experiments to highlight its key properties.

Key facts. Cited by 12 papers in the map (data through 14 Oct 2026). Builds on 28 of its 39 references that are in the map. First posted on arXiv 9 Aug 2021.

Builds on

28 of 39 references.

References · 28

  1. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,513
  2. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,183
  3. 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,897
  4. 2017Prototypical Networks for Few-shot Learningcited by 3,015
  5. 2016Matching Networks for One Shot Learningcited by 2,255
  6. 2019Contrastive Multiview Codingcited by 1,301
  7. 2017Learning to Compare: Relation Network for Few-Shot Learningcited by 1,143
  8. 2016HyperNetworkscited by 1,027
  9. 2018On First-Order Meta-Learning Algorithmscited by 1,018
  10. 2019A Closer Look at Few-shot Classificationcited by 684
  11. 2019Self-labelling via simultaneous clustering and representation learningcited by 528
  12. 2018Meta-Learning with Latent Embedding Optimizationcited by 476
16 more
  1. 2019Meta-Learning with Differentiable Convex Optimizationcited by 461
  2. 2018Meta-Learning for Semi-Supervised Few-Shot Classificationcited by 456
  3. 2020Rethinking Few-Shot Image Classification: a Good Embedding Is All You Need?cited by 437
  4. 2018TADAM: Task dependent adaptive metric for improved few-shot learningcited by 408
  5. 2018Bilevel Programming for Hyperparameter Optimization and Meta-Learningcited by 405
  6. 2018Meta-learning with differentiable closed-form solverscited by 397
  7. 2019Rapid Learning or Feature Reuse? Towards Understanding the Effectiveness of MAMLcited by 359
  8. 2019Meta-Dataset: A Dataset of Datasets for Learning to Learn from Few Examplescited by 315
  9. 2018Few-Shot Learning via Embedding Adaptation with Set-to-Set Functionscited by 294
  10. 2019Charting the Right Manifold: Manifold Mixup for Few-shot Learningcited by 136
  11. 2017Few-Shot Image Recognition by Predicting Parameters from Activationscited by 122
  12. 2020Few-Shot Classification with Feature Map Reconstruction Networkscited by 116
  13. 2018Unsupervised Learning via Meta-Learningcited by 82
  14. 2020Embedding Propagation: Smoother Manifold for Few-Shot Classificationcited by 80
  15. 2019Few-Shot Learning with Embedded Class Models and Shot-Free Meta Trainingcited by 57
  16. 2020The Advantage of Conditional Meta-Learning for Biased Regularization and Fine-Tuningcited by 40

Built on by

Cited by year

Citing papers per year: 2022: 6; 2023: 6.

12 papers.

  1. 2022Inductive and Transductive Few-Shot Video Classification via Appearance and Temporal Alignmentscited by 23
  2. 2023Hubs and Hyperspheres: Reducing Hubness and Improving Transductive Few-shot Learning with Hyperspherical Embeddingscited by 11
  3. 2022Lessons learned from the NeurIPS 2021 MetaDL challenge: Backbone fine-tuning without episodic meta-learning dominates for few-shot learning image classificationcited by 10
  4. 2022The Curse of Low Task Diversity: On the Failure of Transfer Learning to Outperform MAML and Their Empirical Equivalencecited by 10
  5. 2023Few-shot Action Recognition with Captioning Foundation Modelscited by 9
  6. 2023Is Pre-training Truly Better Than Meta-Learning?cited by 6
  7. 2022Schedule-Robust Online Continual Learningcited by 4
  8. 2023Transfer learning for atomistic simulations using GNNs and kernel mean embeddingscited by 4
  9. 2022PLATINUM: Semi-Supervised Model Agnostic Meta-Learning using Submodular Mutual Informationcited by 3
  10. 2023Modularity in Deep Learning: A Surveycited by 3
  11. 2022Robust Meta-Representation Learning via Global Label Inference and Classificationcited by 2
  12. 2023Compositional Prototypical Networks for Few-Shot Classificationcited by 1