mlsanityData to 14 Oct 2026

BatMan-CLR: Making Few-shots Meta-Learners Resilient Against Label Noise

arXiv 2309.06046 ↗First posted 12 Sep 2023cs.LG cs.AI cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

The negative impact of label noise is well studied in classical supervised learning yet remains an open research question in meta-learning. Meta-learners aim to adapt to unseen learning tasks by learning a good initial model in meta-training and consecutively fine-tuning it according to new tasks during meta-testing. In this paper, we present the first extensive analysis of the impact of varying levels of label noise on the performance of state-of-the-art meta-learners, specifically gradient-based N-way K-shot learners. We show that the accuracy of Reptile, iMAML, and foMAML drops by up to 42% on the Omniglot and CifarFS datasets when meta-training is affected by label noise. To strengthen the resilience against label noise, we propose two sampling techniques, namely manifold (Man) and batch manifold (BatMan), which transform the noisy supervised learners into semi-supervised ones to increase the utility of noisy labels. We first construct manifold samples of N-way 2-contrastive-shot tasks through augmentation, learning the embedding via a contrastive loss in meta-training, and then perform classification through zeroing on the embedding in meta-testing. We show that our approach can effectively mitigate the impact of meta-training label noise. Even with 60% wrong labels batman and man can limit the meta-testing accuracy drop to 2.5, 9.4, 1.1 percent points, respectively, with existing meta-learners across the Omniglot, CifarFS, and MiniImagenet datasets.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 19 of its 25 references that are in the map. First posted on arXiv 12 Sep 2023.

Builds on

19 of 25 references.

References · 19

  1. 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,897
  2. 2018On First-Order Meta-Learning Algorithmscited by 1,018
  3. 2019Learning Representations by Maximizing Mutual Information Across Viewscited by 734
  4. 2020Learning from Noisy Labels with Deep Neural Networks: A Surveycited by 582
  5. 2019Meta-Learning with Implicit Gradientscited by 525
  6. 2018Meta-learning with differentiable closed-form solverscited by 397
  7. 2019Gradient Descent with Early Stopping is Provably Robust to Label Noise for Overparameterized Neural Networkscited by 159
  8. 2021Decoupled Contrastive Learningcited by 143
  9. 2018Unsupervised Learning via Meta-Learningcited by 82
  10. 2018Unsupervised Meta-Learning For Few-Shot Image Classificationcited by 75
  11. 2021Meta-learning with an Adaptive Task Schedulercited by 27
  12. 2020Self-Supervised Prototypical Transfer Learning for Few-Shot Classificationcited by 26
7 more
  1. 2022Few-shot Learning with Noisy Labelscited by 23
  2. 2020RNNP: A Robust Few-Shot Learning Approachcited by 9
  3. 2021MAML is a Noisy Contrastive Learner in Classificationcited by 9
  4. 2022Diversity vs. Recognizability: Human-like generalization in one-shot generative modelscited by 9
  5. 2020A Nested Bi-level Optimization Framework for Robust Few Shot Learningcited by 7
  6. 2022Robust Meta-learning with Sampling Noise and Label Noise via Eigen-Reptilecited by 6
  7. 2022Self-Attention Message Passing for Contrastive Few-Shot Learningcited by 4

Built on by

Not cited yet.