mlsanityData to 14 Oct 2026

Learnable Distribution Calibration for Few-Shot Class-Incremental Learning

arXiv 2210.00232 ↗First posted 1 Oct 2022cs.CVcited by 12

Not on the map. Why?

Abstract (arXiv)

Few-shot class-incremental learning (FSCIL) faces challenges of memorizing old class distributions and estimating new class distributions given few training samples. In this study, we propose a learnable distribution calibration (LDC) approach, with the aim to systematically solve these two challenges using a unified framework. LDC is built upon a parameterized calibration unit (PCU), which initializes biased distributions for all classes based on classifier vectors (memory-free) and a single covariance matrix. The covariance matrix is shared by all classes, so that the memory costs are fixed. During base training, PCU is endowed with the ability to calibrate biased distributions by recurrently updating sampled features under the supervision of real distributions. During incremental learning, PCU recovers distributions for old classes to avoid `forgetting', as well as estimating distributions and augmenting samples for new classes to alleviate `over-fitting' caused by the biased distributions of few-shot samples. LDC is theoretically plausible by formatting a variational inference procedure. It improves FSCIL's flexibility as the training procedure requires no class similarity priori. Experiments on CUB200, CIFAR100, and mini-ImageNet datasets show that LDC outperforms the state-of-the-arts by 4.64%, 1.98%, and 3.97%, respectively. LDC's effectiveness is also validated on few-shot learning scenarios.

Key facts. Cited by 12 papers in the map (data through 14 Oct 2026). Builds on 38 of its 60 references that are in the map. First posted on arXiv 1 Oct 2022.

Builds on

38 of 60 references.

References · 38

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2016Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localizationcited by 6,733
  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. 2016Learning without Forgettingcited by 2,496
  6. 2016iCaRL: Incremental Classifier and Representation Learningcited by 2,333
  7. 2016Matching Networks for One Shot Learningcited by 2,255
  8. 2017Continual Learning with Deep Generative Replaycited by 1,158
  9. 2017Learning to Compare: Relation Network for Few-Shot Learningcited by 1,143
  10. 2017PackNet: Adding Multiple Tasks to a Single Network by Iterative Pruningcited by 898
  11. 2018Riemannian Walk for Incremental Learning: Understanding Forgetting and Intransigencecited by 742
  12. 2019Large Scale Incremental Learningcited by 732
26 more
  1. 2017Lifelong Learning with Dynamically Expandable Networkscited by 694
  2. 2019A Closer Look at Few-shot Classificationcited by 684
  3. 2018Overcoming catastrophic forgetting with hard attention to the taskcited by 681
  4. 2018Learning to Learn without Forgetting by Maximizing Transfer and Minimizing Interferencecited by 521
  5. 2020Class-incremental learning: survey and performance evaluation on image classificationcited by 449
  6. 2018Dynamic Few-Shot Visual Learning without Forgettingcited by 385
  7. 2018Meta-Transfer Learning for Few-Shot Learningcited by 306
  8. 2020Few-Shot Class-Incremental Learningcited by 251
  9. 2018Learning without Memorizingcited by 250
  10. 2021Co2L: Contrastive Continual Learningcited by 244
  11. 2020Mnemonics Training: Multi-Class Incremental Learning without Forgettingcited by 221
  12. 2017Incremental Learning of Object Detectors without Catastrophic Forgettingcited by 208
  13. 2021Few-Shot Incremental Learning with Continually Evolved Classifierscited by 187
  14. 2021Free Lunch for Few-shot Learning: Distribution Calibrationcited by 165
  15. 2021Always Be Dreaming: A New Approach for Data-Free Class-Incremental Learningcited by 131
  16. 2022Forward Compatible Few-Shot Class-Incremental Learningcited by 127
  17. 2021Distilling Causal Effect of Data in Class-Incremental Learningcited by 117
  18. 2020Adversarial Feature Hallucination Networks for Few-Shot Learningcited by 110
  19. 2021Self-Promoted Prototype Refinement for Few-Shot Class-Incremental Learningcited by 96
  20. 2019Meta-Learning of Neural Architectures for Few-Shot Learningcited by 73
  21. 2018Incremental Few-Shot Learning with Attention Attractor Networkscited by 66
  22. 2021Few-Shot Lifelong Learningcited by 59
  23. 2019Few-Shot Learning via Saliency-guided Hallucination of Samplescited by 49
  24. 2020XtarNet: Learning to Extract Task-Adaptive Representation for Incremental Few-Shot Learningcited by 31
  25. 2021Continual Learning on Noisy Data Streams via Self-Purified Replaycited by 27
  26. 2020How to trust unlabeled data? Instance Credibility Inference for Few-Shot Learningcited by 16

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 2; 2024: 5; 2025: 2; 2026: 3.

12 papers.

  1. 2023A Survey on Few-Shot Class-Incremental Learningcited by 48
  2. 2023Multimodal Parameter-Efficient Few-Shot Class Incremental Learningcited by 19
  3. 2024Calibrating Higher-Order Statistics for Few-Shot Class-Incremental Learning with Pre-trained Vision Transformerscited by 9
  4. 2024PASS++: A Dual Bias Reduction Framework for Non-Exemplar Class-Incremental Learningcited by 8
  5. 2025Latest Advancements Towards Catastrophic Forgetting under Data Scarcity: A Comprehensive Survey on Few-Shot Class Incremental Learningcited by 4
  6. 2024Covariance-based Space Regularization for Few-shot Class Incremental Learningcited by 2
  7. 2024Controllable Relation Disentanglement for Few-Shot Class-Incremental Learningcited by 1
  8. 2026Few-shot Class-Incremental Learning via Generative Co-Memory Regularizationcited by 1
  9. 2024Incremental Label Distribution Learning with Scalable Graph Convolutional Networkscited by 0
  10. 2025SAFA-SNN: Sparsity-Aware On-Device Few-Shot Class-Incremental Learning with Fast-Adaptive Structure of Spiking Neural Networkcited by 0
  11. 2026CD^2: Constrained Dataset Distillation for Few-Shot Class-Incremental Learningcited by 0
  12. 2026When Sensing Varies with Contexts: Context Probing for Tactile Few-Shot Class-Incremental Learningcited by 0