mlsanityData to 7 Oct 2026

Bridging Few-Shot Learning and Adaptation: New Challenges of Support-Query Shift

arXiv 2105.11804 ↗First posted 25 May 2021cs.LG cs.CVcited by 9

Not on the map. Why?

Abstract (arXiv)

Few-Shot Learning (FSL) algorithms have made substantial progress in learning novel concepts with just a handful of labelled data. To classify query instances from novel classes encountered at test-time, they only require a support set composed of a few labelled samples. FSL benchmarks commonly assume that those queries come from the same distribution as instances in the support set. However, in a realistic set-ting, data distribution is plausibly subject to change, a situation referred to as Distribution Shift (DS). The present work addresses the new and challenging problem of Few-Shot Learning under Support/Query Shift (FSQS) i.e., when support and query instances are sampled from related but different distributions. Our contributions are the following. First, we release a testbed for FSQS, including datasets, relevant baselines and a protocol for a rigorous and reproducible evaluation. Second, we observe that well-established FSL algorithms unsurprisingly suffer from a considerable drop in accuracy when facing FSQS, stressing the significance of our study. Finally, we show that transductive algorithms can limit the inopportune effect of DS. In particular, we study both the role of Batch-Normalization and Optimal Transport (OT) in aligning distributions, bridging Unsupervised Domain Adaptation with FSL. This results in a new method that efficiently combines OT with the celebrated Prototypical Networks. We bring compelling experiments demonstrating the advantage of our method. Our work opens an exciting line of research by providing a testbed and strong baselines. Our code is available at https://github.com/ebennequin/meta-domain-shift.

Key facts. Cited by 9 papers in the map (data through 7 Oct 2026). Builds on 41 of its 67 references that are in the map. First posted on arXiv 25 May 2021.

Builds on

41 of 67 references.

References · 41

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2015Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shiftcited by 6,979
  3. 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,861
  4. 2017Prototypical Networks for Few-shot Learningcited by 3,001
  5. 2019Benchmarking Neural Network Robustness to Common Corruptions and Perturbationscited by 2,857
  6. 2013Sinkhorn Distances: Lightspeed Computation of Optimal Transportation Distancescited by 2,277
  7. 2016Matching Networks for One Shot Learningcited by 2,246
  8. 2014Unsupervised Domain Adaptation by Backpropagationcited by 1,906
  9. 2016Concrete Problems in AI Safetycited by 1,582
  10. 2015Learning Transferable Features with Deep Adaptation Networkscited by 1,323
  11. 2017Learning to Compare: Relation Network for Few-Shot Learningcited by 1,139
  12. 2020In Search of Lost Domain Generalizationcited by 1,042
29 more
  1. 2018On First-Order Meta-Learning Algorithmscited by 1,003
  2. 2020Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptationcited by 844
  3. 2018LEAF: A Benchmark for Federated Settingscited by 780
  4. 2018Computational Optimal Transportcited by 765
  5. 2019A Closer Look at Few-shot Classificationcited by 679
  6. 2015Optimal Transport for Domain Adaptationcited by 457
  7. 2018Meta-Learning for Semi-Supervised Few-Shot Classificationcited by 456
  8. 2020Improving robustness against common corruptions by covariate shift adaptationcited by 427
  9. 2019Bridging Theory and Algorithm for Domain Adaptationcited by 332
  10. 2019Meta-Dataset: A Dataset of Datasets for Learning to Learn from Few Examplescited by 312
  11. 2019A Baseline for Few-Shot Image Classificationcited by 298
  12. 2017Joint Distribution Optimal Transportation for Domain Adaptationcited by 280
  13. 2017Data Augmentation Generative Adversarial Networkscited by 278
  14. 2017Meta Networkscited by 268
  15. 2020Evaluating Prediction-Time Batch Normalization for Robustness under Covariate Shiftcited by 235
  16. 2018DeepJDOT: Deep Joint Distribution Optimal Transport for Unsupervised Domain Adaptationcited by 212
  17. 2018Learning to Propagate Labels: Transductive Propagation Network for Few-shot Learningcited by 208
  18. 2018Low-Shot Learning from Imaginary Datacited by 181
  19. 2018Partial Adversarial Domain Adaptationcited by 127
  20. 2020Domain Adaptation with Conditional Distribution Matching and Generalized Label Shiftcited by 124
  21. 2020Leveraging the Feature Distribution in Transfer-based Few-Shot Learningcited by 72
  22. 2020Tent: Fully Test-time Adaptation by Entropy Minimizationcited by 59
  23. 2020TaskNorm: Rethinking Batch Normalization for Meta-Learningcited by 50
  24. 2020Unraveling Meta-Learning: Understanding Feature Representations for Few-Shot Taskscited by 47
  25. 2020On Episodes, Prototypical Networks, and Few-shot Learningcited by 46
  26. 2020Transductive Information Maximization For Few-Shot Learningcited by 44
  27. 2020Domain-Adaptive Few-Shot Learningcited by 40
  28. 2020Robust Domain Adaptation: Representations, Weights and Inductive Biascited by 6
  29. 2020Target Consistency for Domain Adaptation: when Robustness meets Transferabilitycited by 1

Built on by

Cited by year

Citing papers per year: 2022: 4; 2023: 3; 2024: 1; 2025: 1.

9 papers.

  1. 2023Open-Set Likelihood Maximization for Few-Shot Learningcited by 10
  2. 2022Towards Practical Few-Shot Query Sets: Transductive Minimum Description Length Inferencecited by 9
  3. 2022Model-Agnostic Few-Shot Open-Set Recognitioncited by 2
  4. 2022PGADA: Perturbation-Guided Adversarial Alignment for Few-shot Learning Under the Support-Query Shiftcited by 1
  5. 2023Dual Adversarial Alignment for Realistic Support-Query Shift Few-shot Learningcited by 1
  6. 2025CCoMAML: Efficient Cattle Identification Using Cooperative Model-Agnostic Meta-Learningcited by 1
  7. 2022Constrained Few-Shot Learning: Human-Like Low Sample Complexity Learning and Non-Episodic Text Classificationcited by 0
  8. 2023Unsupervised Representation Learning to Aid Semi-Supervised Meta Learningcited by 0
  9. 2024Few and Fewer: Learning Better from Few Examples Using Fewer Base Classescited by 0