mlsanityData to 7 Oct 2026

A Survey on Few-Shot Class-Incremental Learning

arXiv 2304.08130 ↗First posted 17 Apr 2023cs.CVcited by 48

Not on the map. Why?

Abstract (arXiv)

Large deep learning models are impressive, but they struggle when real-time data is not available. Few-shot class-incremental learning (FSCIL) poses a significant challenge for deep neural networks to learn new tasks from just a few labeled samples without forgetting the previously learned ones. This setup easily leads to catastrophic forgetting and overfitting problems, severely affecting model performance. Studying FSCIL helps overcome deep learning model limitations on data volume and acquisition time, while improving practicality and adaptability of machine learning models. This paper provides a comprehensive survey on FSCIL. Unlike previous surveys, we aim to synthesize few-shot learning and incremental learning, focusing on introducing FSCIL from two perspectives, while reviewing over 30 theoretical research studies and more than 20 applied research studies. From the theoretical perspective, we provide a novel categorization approach that divides the field into five subcategories, including traditional machine learning methods, meta-learning based methods, feature and feature space-based methods, replay-based methods, and dynamic network structure-based methods. We also evaluate the performance of recent theoretical research on benchmark datasets of FSCIL. From the application perspective, FSCIL has achieved impressive achievements in various fields of computer vision such as image classification, object detection, and image segmentation, as well as in natural language processing and graph. We summarize the important applications. Finally, we point out potential future research directions, including applications, problem setups, and theory development. Overall, this paper offers a comprehensive analysis of the latest advances in FSCIL from a methodological, performance, and application perspective.

Key facts. Cited by 48 papers in the map (data through 7 Oct 2026). Builds on 99 of its 168 references that are in the map. First posted on arXiv 17 Apr 2023.

Builds on

99 of 168 references.

References · 99

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
  3. 2014Microsoft COCO: Common Objects in Contextcited by 17,498
  4. 2017Mask R-CNNcited by 7,259
  5. 2017Inductive Representation Learning on Large Graphscited by 6,284
  6. 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,861
  7. 2020Deformable DETR: Deformable Transformers for End-to-End Object Detectioncited by 3,410
  8. 2017Prototypical Networks for Few-shot Learningcited by 3,001
  9. 2016iCaRL: Incremental Classifier and Representation Learningcited by 2,313
  10. 2015Character-level Convolutional Networks for Text Classificationcited by 2,293
  11. 2016Matching Networks for One Shot Learningcited by 2,246
  12. 2016Progressive Neural Networkscited by 1,332
38 more
  1. 2018Continual Lifelong Learning with Neural Networks: A Reviewcited by 1,313
  2. 2019LVIS: A Dataset for Large Vocabulary Instance Segmentationcited by 1,215
  3. 2019Objects as Pointscited by 1,172
  4. 2019A Comprehensive Survey on Transfer Learningcited by 1,154
  5. 2020Meta-Learning in Neural Networks: A Surveycited by 944
  6. 2019Relational Knowledge Distillationcited by 873
  7. 2016COCO-Stuff: Thing and Stuff Classes in Contextcited by 858
  8. 2016Abstractive Text Summarization Using Sequence-to-Sequence RNNs and Beyondcited by 809
  9. 2019Large Scale Incremental Learningcited by 727
  10. 2017Dynamic Routing Between Capsulescited by 689
  11. 2022DualPrompt: Complementary Prompting for Rehearsal-free Continual Learningcited by 635
  12. 2019GraphSAINT: Graph Sampling Based Inductive Learning Methodcited by 618
  13. 2019Three scenarios for continual learningcited by 610
  14. 2018End-to-End Incremental Learningcited by 596
  15. 2018Learning to Learn without Forgetting by Maximizing Transfer and Minimizing Interferencecited by 511
  16. 2018Meta-Learning with Latent Embedding Optimizationcited by 476
  17. 2019Generalizing from a Few Examples: A Survey on Few-Shot Learningcited by 462
  18. 2018Meta-Learning for Semi-Supervised Few-Shot Classificationcited by 456
  19. 2018Dynamic Few-Shot Visual Learning without Forgettingcited by 385
  20. 2016Expert Gate: Lifelong Learning with a Network of Expertscited by 369
  21. 2019Dynamic Convolution: Attention over Convolution Kernelscited by 355
  22. 2017One-Shot Learning for Semantic Segmentationcited by 337
  23. 2017Deep Gaussian Embedding of Graphs: Unsupervised Inductive Learning via Rankingcited by 332
  24. 2017Meta Networkscited by 268
  25. 2018Few-shot Object Detection via Feature Reweightingcited by 265
  26. 2018FewRel: A Large-Scale Supervised Few-Shot Relation Classification Dataset with State-of-the-Art Evaluationcited by 254
  27. 2020Few-Shot Class-Incremental Learningcited by 249
  28. 2022FOSTER: Feature Boosting and Compression for Class-Incremental Learningcited by 228
  29. 2023Class-Incremental Learning: A Surveycited by 209
  30. 2017Incremental Learning of Object Detectors without Catastrophic Forgettingcited by 208
  31. 2021Few-Shot Incremental Learning with Continually Evolved Classifierscited by 184
  32. 2018IDD: A Dataset for Exploring Problems of Autonomous Navigation in Unconstrained Environmentscited by 163
  33. 2022A Comprehensive Survey of Few-shot Learning: Evolution, Applications, Challenges, and Opportunitiescited by 162
  34. 2018Task-Agnostic Meta-Learning for Few-shot Learningcited by 151
  35. 2018Bounding Box Regression with Uncertainty for Accurate Object Detectioncited by 143
  36. 2015A Large-Scale Car Dataset for Fine-Grained Categorization and Verificationcited by 129
  37. 2022Forward Compatible Few-Shot Class-Incremental Learningcited by 125
  38. 2021Overcoming Catastrophic Forgetting in Incremental Few-Shot Learning by Finding Flat Minimacited by 118

50 most cited of 99 All 99 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 1; 2024: 24; 2025: 19; 2026: 4.

48 papers.

Sort:

By area:

  1. 2024Pre-trained Vision and Language Transformers Are Few-Shot Incremental Learnerscited by 35
  2. 2024OrCo: Towards Better Generalization via Orthogonality and Contrast for Few-Shot Class-Incremental Learningcited by 25
  3. 2024Few-Shot Class Incremental Learning with Attention-Aware Self-Adaptive Promptcited by 15
  4. 2024Open World Object Detection: A Surveycited by 13
  5. 2024Learning Prompt with Distribution-Based Feature Replay for Few-Shot Class-Incremental Learningcited by 12
  6. 2024Cross-border Commodity Pricing Strategy Optimization via Mixed Neural Network for Time Series Analysiscited by 12
  7. 2024Calibrating Higher-Order Statistics for Few-Shot Class-Incremental Learning with Pre-trained Vision Transformerscited by 9
  8. 2024Real-time Monitoring of Lower Limb Movement Resistance Based on Deep Learningcited by 8
  9. 2024A Survey of Few-Shot Learning for Biomedical Time Seriescited by 7
  10. 2024Mamba-FSCIL: Dynamic Adaptation with Selective State Space Model for Few-Shot Class-Incremental Learningcited by 7
  11. 2024Towards Incremental Learning in Large Language Models: A Critical Reviewcited by 6
  12. 2024MAPLE: A Framework for Active Preference Learning Guided by Large Language Modelscited by 6
36 more
  1. 2024Hyperdimensional Representation Learning for Node Classification and Link Predictioncited by 5
  2. 2024Towards Few-Shot Learning in the Open World: A Review and Beyondcited by 5
  3. 2024MSDNet: Multi-Scale Decoder for Few-Shot Semantic Segmentation via Transformer-Guided Prototypingcited by 5
  4. 2025Adaptive Decision Boundary for Few-Shot Class-Incremental Learningcited by 5
  5. 2024Enhancing Few-Shot Image Classification through Learnable Multi-Scale Embedding and Attention Mechanismscited by 4
  6. 2024CFTS-GAN: Continual Few-Shot Teacher Student for Generative Adversarial Networkscited by 4
  7. 2025Latest Advancements Towards Catastrophic Forgetting under Data Scarcity: A Comprehensive Survey on Few-Shot Class Incremental Learningcited by 3
  8. 2025Global Convergence of Continual Learning on Non-IID Datacited by 3
  9. 2025ProtoConNet: Prototypical Augmentation and Alignment for Open-Set Few-Shot Image Classificationcited by 3
  10. 2024Covariance-based Space Regularization for Few-shot Class Incremental Learningcited by 2
  11. 2024Adapter-Enhanced Semantic Prompting for Continual Learningcited by 2
  12. 2025Beyond Single-Sentence Prompts: Upgrading Value Alignment Benchmarks with Dialogues and Storiescited by 2
  13. 2025Can Synthetic Images Conquer Forgetting? Beyond Unexplored Doubts in Few-Shot Class-Incremental Learningcited by 2
  14. 2024PL-FSCIL: Harnessing the Power of Prompts for Few-Shot Class-Incremental Learningcited by 1
  15. 2024Dental Severity Assessment through Few-shot Learning and SBERT Fine-tuningcited by 1
  16. 2024Consistency-Guided Asynchronous Contrastive Tuning for Few-Shot Class-Incremental Tuning of Foundation Modelscited by 1
  17. 2024TEAL: New Selection Strategy for Small Buffers in Experience Replay Class Incremental Learningcited by 1
  18. 2024From Uncertainty to Clarity: Uncertainty-Guided Class-Incremental Learning for Limited Biomedical Samples via Semantic Expansioncited by 1
  19. 2025Controllable Forgetting Mechanism for Few-Shot Class-Incremental Learningcited by 1
  20. 2025Does Prior Data Matter? Exploring Joint Training in the Context of Few-Shot Class-Incremental Learningcited by 1
  21. 2025CoLoR-GAN: Continual Few-Shot Learning with Low-Rank Adaptation in Generative Adversarial Networkscited by 1
  22. 2023A Forward and Backward Compatible Framework for Few-shot Class-incremental Pill Recognitioncited by 0
  23. 2025Brain-inspired analogical mixture prototypes for few-shot class-incremental learningcited by 0
  24. 2025An experimental approach on Few Shot Class Incremental Learningcited by 0
  25. 2025Adaptive Additive Parameter Updates of Vision Transformers for Few-Shot Continual Learningcited by 0
  26. 2025MEGA: Second-Order Gradient Alignment for Catastrophic Forgetting Mitigation in GFSCILcited by 0
  27. 2025Class-Incremental Learning for Honey Botanical Origin Classification with Hyperspectral Images: A Study with Continual Backpropagationcited by 0
  28. 2025Sculpting Margin Penalty: Intra-Task Adapter Merging and Classifier Calibration for Few-Shot Class-Incremental Learningcited by 0
  29. 2025Exploring the Tradeoff Between Diversity and Discrimination for Continuous Category Discoverycited by 0
  30. 2025MoTiC: Momentum Tightness and Contrast for Few-Shot Class-Incremental Learningcited by 0
  31. 2025"Stop replacing salt with sugar!'': Towards Intuitive Human-Agent Teachingcited by 0
  32. 2025Diffusion-Classifier Synergy: Reward-Aligned Learning via Mutual Boosting Loop for FSCILcited by 0
  33. 2026PKI: Prior Knowledge-Infused Neural Network for Few-Shot Class-Incremental Learningcited by 0
  34. 2026A Predict-then-Correct Loop Based on Few-Shot Continuous Contextual Bandit for Demand Forecastingcited by 0
  35. 2026A Kernel-Based Modular Discriminant Analysis Framework for Small-Sample Learningcited by 0
  36. 2026Decoupled and Distilled: Task-Adaptive LoRA-Teachers with Ensemble Knowledge Transfer for Few-Shot Class-Incremental Learningcited by 0