Few-shot Class-incremental Learning: A Survey

Few-shot Class-Incremental Learning (FSCIL) presents a unique challenge in machine learning, as it necessitates the continuous learning of new classes from sparse labeled training samples without forgetting previous knowledge. While this field has seen recent progress, it remains an active area of exploration. This paper aims to provide a comprehensive and systematic review of FSCIL. In our in-depth examination, we delve into various facets of FSCIL, encompassing the problem definition, the discussion of primary challenges of unreliable empirical risk minimization and the stability-plasticity dilemma, general schemes, and relevant problems of incremental learning and few-shot learning. Besides, we offer an overview of benchmark datasets and evaluation metrics. Furthermore, we introduce the classification methods in FSCIL from data-based, structure-based, and optimization-based approaches and the object detection methods in FSCIL from anchor-free and anchor-based approaches. Beyond these, we illuminate several promising research directions within FSCIL that merit further investigation.

[1]  S. Biswas,et al.  S3C: Self-Supervised Stochastic Classifiers for Few-Shot Class-Incremental Learning , 2023, ECCV.

[2]  Zhanzhan Cheng,et al.  Few-Shot Class-Incremental Learning via Class-Aware Bilateral Distillation , 2023, 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[3]  Dewen Hu,et al.  Few-shot Class-incremental Pill Recognition , 2023, ArXiv.

[4]  Weijun Li,et al.  A Survey on Few-Shot Class-Incremental Learning , 2023, ArXiv.

[5]  Ross B. Girshick,et al.  Segment Anything , 2023, 2023 IEEE/CVF International Conference on Computer Vision (ICCV).

[6]  Qianhua He,et al.  Few-shot class-incremental audio classification via discriminative prototype learning , 2023, Expert Syst. Appl..

[7]  Vishal M. Patel,et al.  Deep Learning for Cross-Domain Few-Shot Visual Recognition: A Survey , 2023, ArXiv.

[8]  Han-Jia Ye,et al.  Deep Class-Incremental Learning: A Survey , 2023, ArXiv.

[9]  Philip H. S. Torr,et al.  Neural Collapse Inspired Feature-Classifier Alignment for Few-Shot Class Incremental Learning , 2023, ICLR.

[10]  Jun Zhu,et al.  A Comprehensive Survey of Continual Learning: Theory, Method and Application , 2023, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[11]  Yawen Cui,et al.  Uncertainty-Aware Distillation for Semi-Supervised Few-Shot Class-Incremental Learning , 2023, IEEE transactions on neural networks and learning systems.

[12]  Yongsheng Gao,et al.  SSFE-Net: Self-Supervised Feature Enhancement for Ultra-Fine-Grained Few-Shot Class Incremental Learning , 2023, IEEE Workshop/Winter Conference on Applications of Computer Vision.

[13]  T. Tuytelaars,et al.  Three types of incremental learning , 2022, Nat. Mac. Intell..

[14]  Shanghang Zhang,et al.  Margin-Based Few-Shot Class-Incremental Learning with Class-Level Overfitting Mitigation , 2022, NeurIPS.

[15]  S. Chaudhuri,et al.  Semantics-Driven Generative Replay for Few-Shot Class Incremental Learning , 2022, ACM Multimedia.

[16]  Guangle Yao,et al.  Feature Distribution Distillation-Based Few Shot Class Incremental Learning , 2022, 2022 5th International Conference on Pattern Recognition and Artificial Intelligence (PRAI).

[17]  Zhong Ji,et al.  Memorizing Complementation Network for Few-Shot Class-Incremental Learning , 2022, IEEE Transactions on Image Processing.

[18]  B. Lovell,et al.  Few-Shot Class-Incremental Learning from an Open-Set Perspective , 2022, ECCV.

[19]  Yang Wang,et al.  Few-Shot Class-Incremental Learning via Entropy-Regularized Data-Free Replay , 2022, ECCV.

[20]  Jun Li,et al.  Few-shot class-incremental learning based on representation enhancement , 2022, Journal of Electronic Imaging.

[21]  Şerif Bahtiyar,et al.  Data poisoning attacks against machine learning algorithms , 2022, Expert Syst. Appl..

[22]  T. Boult,et al.  Variable Few Shot Class Incremental and Open World Learning , 2022, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW).

[23]  T. Boult,et al.  Few-Shot Class Incremental Learning Leveraging Self-Supervised Features , 2022, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW).

[24]  Yang Wang,et al.  MetaFSCIL: A Meta-Learning Approach for Few-Shot Class Incremental Learning , 2022, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[25]  S. Todorovic,et al.  iFS-RCNN: An Incremental Few-shot Instance Segmenter , 2022, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[26]  Qixiang Ye,et al.  Dynamic Support Network for Few-Shot Class Incremental Learning , 2022, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[27]  Yisheng Song,et al.  A Comprehensive Survey of Few-shot Learning: Evolution, Applications, Challenges, and Opportunities , 2022, ACM Comput. Surv..

[28]  Yongqiang Zhang,et al.  Incremental-DETR: Incremental Few-Shot Object Detection via Self-Supervised Learning , 2022, AAAI.

[29]  Hanli Wang,et al.  Meta-Learning-Based Incremental Few-Shot Object Detection , 2022, IEEE Transactions on Circuits and Systems for Video Technology.

[30]  De-Chuan Zhan,et al.  Few-Shot Class-Incremental Learning by Sampling Multi-Phase Tasks , 2022, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[31]  L. Benini,et al.  Constrained Few-shot Class-incremental Learning , 2022, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[32]  Kevin J Liang,et al.  Sylph: A Hypernetwork Framework for Incremental Few-shot Object Detection , 2022, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[33]  Fu Lee Wang,et al.  Forward Compatible Few-Shot Class-Incremental Learning , 2022, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[34]  Shafiq R. Joty,et al.  Continual Few-shot Relation Learning via Embedding Space Regularization and Data Augmentation , 2022, ACL.

[35]  Xu Yang,et al.  Incremental few-shot object detection via knowledge transfer , 2022, Pattern Recognit. Lett..

[36]  Xi Li,et al.  MgSvF: Multi-Grained Slow versus Fast Framework for Few-Shot Class-Incremental Learning , 2020, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[37]  Aidong Zhang,et al.  Few-Shot Class-Incremental Learning with Meta-Learned Class Structures , 2021, 2021 International Conference on Data Mining Workshops (ICDMW).

[38]  Xiao-Ming Wu,et al.  Overcoming Catastrophic Forgetting in Incremental Few-Shot Learning by Finding Flat Minima , 2021, NeurIPS.

[39]  Issam H. Laradji,et al.  A Survey of Self-Supervised and Few-Shot Object Detection , 2021, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[40]  Mehrtash Harandi,et al.  Synthesized Feature based Few-Shot Class-Incremental Learning on a Mixture of Subspaces , 2021, 2021 IEEE/CVF International Conference on Computer Vision (ICCV).

[41]  Mohammad Tavakolian,et al.  Semi-Supervised Few-Shot Class-Incremental Learning , 2021, 2021 IEEE International Conference on Image Processing (ICIP).

[42]  Bernt Schiele,et al.  Generalized and Incremental Few-Shot Learning by Explicit Learning and Calibration without Forgetting , 2021, 2021 IEEE/CVF International Conference on Computer Vision (ICCV).

[43]  Wendy Hall,et al.  What Can Knowledge Bring to Machine Learning?—A Survey of Low-shot Learning for Structured Data , 2021, ACM Trans. Intell. Syst. Technol..

[44]  Yu Wang,et al.  Few-Shot Continual Learning for Audio Classification , 2021, ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[45]  Zheng-Jun Zha,et al.  Self-Promoted Prototype Refinement for Few-Shot Class-Incremental Learning , 2021, 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[46]  Yihong Gong,et al.  Few-Shot Class-Incremental Learning via Relation Knowledge Distillation , 2021, AAAI.

[47]  Xiaoxu Li,et al.  Deep metric learning for few-shot image classification: A Review of recent developments , 2021, Pattern Recognit..

[48]  Ronald Poppe,et al.  Incremental Few-Shot Instance Segmentation , 2021, 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[49]  Yinghui Xu,et al.  Few-Shot Incremental Learning with Continually Evolved Classifiers , 2021, 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[50]  Qixiang Ye,et al.  Learnable Expansion-and-Compression Network for Few-shot Class-Incremental Learning , 2021, ArXiv.

[51]  M. Harandi,et al.  Semantic-aware Knowledge Distillation for Few-Shot Class-Incremental Learning , 2021, 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[52]  Piyush Rai,et al.  Few-Shot Lifelong Learning , 2021, AAAI.

[53]  Ilya Sutskever,et al.  Learning Transferable Visual Models From Natural Language Supervision , 2021, ICML.

[54]  Hyunwoo J. Kim,et al.  Online Continual Learning in Image Classification: An Empirical Survey , 2021, Neurocomputing.

[55]  Ioannis Kanellos,et al.  A Comprehensive Study of Class Incremental Learning Algorithms for Visual Tasks , 2020, Neural Networks.

[56]  Joost van de Weijer,et al.  Class-Incremental Learning: Survey and Performance Evaluation on Image Classification , 2020, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[57]  Pinghua Gong,et al.  Learning from Very Few Samples: A Survey , 2020, ArXiv.

[58]  David L. Donoho,et al.  Prevalence of neural collapse during the terminal phase of deep learning training , 2020, Proceedings of the National Academy of Sciences.

[59]  Jianping Gou,et al.  Knowledge Distillation: A Survey , 2020, International Journal of Computer Vision.

[60]  Nicolas Usunier,et al.  End-to-End Object Detection with Transformers , 2020, ECCV.

[61]  Zhi-Hua Zhou,et al.  Heterogeneous Few-Shot Model Rectification With Semantic Mapping , 2020, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[62]  Xiaopeng Hong,et al.  Few-Shot Class-Incremental Learning , 2020, 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[63]  Rama Chellappa,et al.  The Devil is in the Details: Self-Supervised Attention for Vehicle Re-Identification , 2020, ECCV.

[64]  Timothy M. Hospedales,et al.  Meta-Learning in Neural Networks: A Survey , 2020, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[65]  Joost van de Weijer,et al.  Semantic Drift Compensation for Class-Incremental Learning , 2020, 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[66]  Jaekyun Moon,et al.  XtarNet: Learning to Extract Task-Adaptive Representation for Incremental Few-Shot Learning , 2020, ICML.

[67]  Trevor Darrell,et al.  Frustratingly Simple Few-Shot Object Detection , 2020, ICML.

[68]  Tao Xiang,et al.  Incremental Few-Shot Object Detection , 2020, 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[69]  Shifeng Zhang,et al.  Bridging the Gap Between Anchor-Based and Anchor-Free Detection via Adaptive Training Sample Selection , 2019, 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[70]  Sung Ju Hwang,et al.  Self-supervised Label Augmentation via Input Transformations , 2019, ICML.

[71]  Tin Lay Nwe,et al.  Meta Module Generation for Fast Few-Shot Incremental Learning , 2019, 2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW).

[72]  Tinne Tuytelaars,et al.  A Continual Learning Survey: Defying Forgetting in Classification Tasks , 2019, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[73]  Jiliang Tang,et al.  Adversarial Attacks and Defenses in Images, Graphs and Text: A Review , 2019, International Journal of Automation and Computing.

[74]  Hasan Şakir Bilge,et al.  Deep Metric Learning: A Survey , 2019, Symmetry.

[75]  Xingyi Zhou,et al.  Objects as Points , 2019, ArXiv.

[76]  James T. Kwok,et al.  Generalizing from a Few Examples , 2019, ACM Comput. Surv..

[77]  Yu-Chiang Frank Wang,et al.  A Closer Look at Few-shot Classification , 2019, ICLR.

[78]  Hao Chen,et al.  FCOS: Fully Convolutional One-Stage Object Detection , 2019, 2019 IEEE/CVF International Conference on Computer Vision (ICCV).

[79]  Renjie Liao,et al.  Incremental Few-Shot Learning with Attention Attractor Networks , 2018, NeurIPS.

[80]  Davide Maltoni,et al.  Continuous Learning in Single-Incremental-Task Scenarios , 2018, Neural Networks.

[81]  Yarin Gal,et al.  Towards Robust Evaluations of Continual Learning , 2018, ArXiv.

[82]  Nikos Komodakis,et al.  Dynamic Few-Shot Visual Learning Without Forgetting , 2018, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[83]  Stefan Wermter,et al.  Continual Lifelong Learning with Neural Networks: A Review , 2018, Neural Networks.

[84]  Matthew A. Brown,et al.  Low-Shot Learning with Imprinted Weights , 2017, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[85]  Svetlana Lazebnik,et al.  PackNet: Adding Multiple Tasks to a Single Network by Iterative Pruning , 2017, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[86]  Léon Bottou,et al.  Wasserstein Generative Adversarial Networks , 2017, ICML.

[87]  Lukasz Kaiser,et al.  Attention is All you Need , 2017, NIPS.

[88]  Ross B. Girshick,et al.  Mask R-CNN , 2017, 1703.06870.

[89]  Sergey Levine,et al.  Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks , 2017, ICML.

[90]  Christoph H. Lampert,et al.  iCaRL: Incremental Classifier and Representation Learning , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[91]  Jorge Nocedal,et al.  Optimization Methods for Large-Scale Machine Learning , 2016, SIAM Rev..

[92]  Oriol Vinyals,et al.  Matching Networks for One Shot Learning , 2016, NIPS.

[93]  Wei-Lun Chao,et al.  An Empirical Study and Analysis of Generalized Zero-Shot Learning for Object Recognition in the Wild , 2016, ECCV.

[94]  Kaiming He,et al.  Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks , 2015, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[95]  Michael S. Bernstein,et al.  ImageNet Large Scale Visual Recognition Challenge , 2014, International Journal of Computer Vision.

[96]  Pietro Perona,et al.  Microsoft COCO: Common Objects in Context , 2014, ECCV.

[97]  Yoshua Bengio,et al.  An Empirical Investigation of Catastrophic Forgeting in Gradient-Based Neural Networks , 2013, ICLR.

[98]  Koen E. A. van de Sande,et al.  Selective Search for Object Recognition , 2013, International Journal of Computer Vision.

[99]  Pascal Vincent,et al.  Representation Learning: A Review and New Perspectives , 2012, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[100]  Pietro Perona,et al.  The Caltech-UCSD Birds-200-2011 Dataset , 2011 .

[101]  Christopher K. I. Williams,et al.  International Journal of Computer Vision manuscript No. (will be inserted by the editor) The PASCAL Visual Object Classes (VOC) Challenge , 2022 .

[102]  Léon Bottou,et al.  The Tradeoffs of Large Scale Learning , 2007, NIPS.

[103]  T. Martínez,et al.  Competitive Hebbian Learning Rule Forms Perfectly Topology Preserving Maps , 1993 .

[104]  David J. Spiegelhalter,et al.  Sequential updating of conditional probabilities on directed graphical structures , 1990, Networks.

[105]  Dong-Jun Han,et al.  Warping the Space: Weight Space Rotation for Class-Incremental Few-Shot Learning , 2023, ICLR.

[106]  Feng Zhou,et al.  Few-Shot Class-Incremental SAR Target Recognition Based on Hierarchical Embedding and Incremental Evolutionary Network , 2023, IEEE Transactions on Geoscience and Remote Sensing.

[107]  Xicheng Lu,et al.  Decoupled Two-Phase Framework for Class-Incremental Few-Shot Named Entity Recognition , 2023, Tsinghua Science and Technology.

[108]  Gangyao Kuang,et al.  Few-Shot Class-Incremental SAR Target Recognition via Cosine Prototype Learning , 2023, IEEE Transactions on Geoscience and Remote Sensing.

[109]  M. Pietikainen,et al.  Uncertainty-Guided Semi-Supervised Few-Shot Class-Incremental Learning With Knowledge Distillation , 2023, IEEE Transactions on Multimedia.

[110]  Ricardo Henao,et al.  Few-Shot Class-Incremental Learning for Named Entity Recognition , 2022, ACL.

[111]  Horst-Michael Groß,et al.  Few-Shot Object Detection: A Survey , 2021, ArXiv.

[112]  Koichiro Yamauchi,et al.  Few-Shot Class Incremental Learning with Generative Feature Replay , 2021, ICPRAM.

[113]  Alex Krizhevsky,et al.  Learning Multiple Layers of Features from Tiny Images , 2009 .

[114]  Radford M. Neal.  Bayesian learning for neural networks , 1995 .