mlsanityData to 14 Oct 2026

Coarse-To-Fine Incremental Few-Shot Learning

arXiv 2111.14806 ↗First posted 24 Nov 2021cs.CV cs.LGcited by 9

Not on the map. Why?

Abstract (arXiv)

Different from fine-tuning models pre-trained on a large-scale dataset of preset classes, class-incremental learning (CIL) aims to recognize novel classes over time without forgetting pre-trained classes. However, a given model will be challenged by test images with finer-grained classes, e.g., a basenji is at most recognized as a dog. Such images form a new training set (i.e., support set) so that the incremental model is hoped to recognize a basenji (i.e., query) as a basenji next time. This paper formulates such a hybrid natural problem of coarse-to-fine few-shot (C2FS) recognition as a CIL problem named C2FSCIL, and proposes a simple, effective, and theoretically-sound strategy Knowe: to learn, normalize, and freeze a classifier's weights from fine labels, once learning an embedding space contrastively from coarse labels. Besides, as CIL aims at a stability-plasticity balance, new overall performance metrics are proposed. In that sense, on CIFAR-100, BREEDS, and tieredImageNet, Knowe outperforms all recent relevant CIL/FSCIL methods that are tailored to the new problem setting for the first time.

Key facts. Cited by 9 papers in the map (data through 14 Oct 2026). Builds on 45 of its 66 references that are in the map. First posted on arXiv 24 Nov 2021.

Builds on

45 of 66 references.

References · 45

  1. 2015Distilling the Knowledge in a Neural Networkcited by 10,324
  2. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,513
  3. 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,897
  4. 2016Overcoming catastrophic forgetting in neural networkscited by 4,787
  5. 2017Prototypical Networks for Few-shot Learningcited by 3,015
  6. 2016Learning without Forgettingcited by 2,496
  7. 2016iCaRL: Incremental Classifier and Representation Learningcited by 2,333
  8. 2017PackNet: Adding Multiple Tasks to a Single Network by Iterative Pruningcited by 898
  9. 2018Learning to Learn without Forgetting by Maximizing Transfer and Minimizing Interferencecited by 521
  10. 2018Meta-Learning for Semi-Supervised Few-Shot Classificationcited by 456
  11. 2020Class-incremental learning: survey and performance evaluation on image classificationcited by 449
  12. 2020Rethinking Few-Shot Image Classification: a Good Embedding Is All You Need?cited by 437
33 more
  1. 2018Dynamic Few-Shot Visual Learning without Forgettingcited by 385
  2. 2019Maintaining Discrimination and Fairness in Class Incremental Learningcited by 283
  3. 2020Few-Shot Class-Incremental Learningcited by 251
  4. 2018Learning without Memorizingcited by 250
  5. 2021Online Continual Learning in Image Classification: An Empirical Surveycited by 245
  6. 2021Co2L: Contrastive Continual Learningcited by 244
  7. 2017Overcoming Catastrophic Forgetting by Incremental Moment Matchingcited by 243
  8. 2017NormFace: L2 Hypersphere Embedding for Face Verificationcited by 237
  9. 2014Towards Open World Recognitioncited by 226
  10. 2020Mnemonics Training: Multi-Class Incremental Learning without Forgettingcited by 221
  11. 2019Compacting, Picking and Growing for Unforgetting Continual Learningcited by 207
  12. 2021Few-Shot Incremental Learning with Continually Evolved Classifierscited by 187
  13. 2020BREEDS: Benchmarks for Subpopulation Shiftcited by 155
  14. 2020Adaptive Aggregation Networks for Class-Incremental Learningcited by 149
  15. 2021Training Networks in Null Space of Feature Covariance for Continual Learningcited by 143
  16. 2021Distilling Causal Effect of Data in Class-Incremental Learningcited by 117
  17. 2020Continual Learning with Node-Importance based Adaptive Group Sparse Regularizationcited by 105
  18. 2020Continual Deep Learning by Functional Regularisation of Memorable Pastcited by 105
  19. 2018Incremental Few-Shot Learning with Attention Attractor Networkscited by 66
  20. 2021Subspace Regularizers for Few-Shot Class Incremental Learningcited by 56
  21. 2020ScaIL: Classifier Weights Scaling for Class Incremental Learningcited by 47
  22. 2020Continual Learning with Extended Kronecker-factored Approximate Curvaturecited by 39
  23. 2021Generalized and Incremental Few-Shot Learning by Explicit Learning and Calibration without Forgettingcited by 33
  24. 2021Continual Learning via Bit-Level Information Preservingcited by 32
  25. 2020Fine-grained Angular Contrastive Learning with Coarse Labelscited by 30
  26. 2020Learning from Few Samples: A Surveycited by 23
  27. 2021Efficient Algorithms for Learning from Coarse Labelscited by 21
  28. 2020IIRC: Incremental Implicitly-Refined Classificationcited by 20
  29. 2018Small Sample Learning in Big Data Eracited by 18
  30. 2021Rectification-based Knowledge Retention for Continual Learningcited by 15
  31. 2020Weakly Supervised Representation Learning with Coarse Labelscited by 11
  32. 2020Initial Classifier Weights Replay for Memoryless Class Incremental Learningcited by 11
  33. 2020Towards Cross-Granularity Few-Shot Learning: Coarse-to-Fine Pseudo-Labeling with Visual-Semantic Meta-Embeddingcited by 7

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2022: 1; 2023: 2; 2024: 1; 2025: 4; 2026: 1.

9 papers.

  1. 2024Continual Adapter Tuning with Semantic Shift Compensation for Class-Incremental Learningcited by 10
  2. 2023Continual Learning for LiDAR Semantic Segmentation: Class-Incremental and Coarse-to-Fine strategies on Sparse Datacited by 9
  3. 2023Hyperbolic Space with Hierarchical Margin Boosts Fine-Grained Learning from Coarse Labelscited by 8
  4. 2025StPR: Spatiotemporal Preservation and Routing for Exemplar-Free Video Class-Incremental Learningcited by 7
  5. 2022Long-Tailed Classification with Gradual Balanced Loss and Adaptive Feature Generationcited by 2
  6. 2025OpenEarthSensing: Large-Scale Fine-Grained Benchmark for Open-World Remote Sensingcited by 2
  7. 2025PTMs-TSCIL Pre-Trained Models Based Class-Incremental Learningcited by 1
  8. 2025Hyperbolic Coarse-to-Fine Few-Shot Class-Incremental Learningcited by 0
  9. 2026Let's Split Up: Zero-Shot Classifier Edits for Fine-Grained Video Understandingcited by 0