Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
110 tokens/sec
GPT-4o
56 tokens/sec
Gemini 2.5 Pro Pro
44 tokens/sec
o3 Pro
6 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Enhanced Few-Shot Class-Incremental Learning via Ensemble Models (2401.07208v2)

Published 14 Jan 2024 in cs.CV

Abstract: Few-shot class-incremental learning (FSCIL) aims to continually fit new classes with limited training data, while maintaining the performance of previously learned classes. The main challenges are overfitting the rare new training samples and forgetting old classes. While catastrophic forgetting has been extensively studied, the overfitting problem has attracted less attention in FSCIL. To tackle overfitting challenge, we design a new ensemble model framework cooperated with data augmentation to boost generalization. In this way, the enhanced model works as a library storing abundant features to guarantee fast adaptation to downstream tasks. Specifically, the multi-input multi-output ensemble structure is applied with a spatial-aware data augmentation strategy, aiming at diversifying the feature extractor and alleviating overfitting in incremental sessions. Moreover, self-supervised learning is also integrated to further improve the model generalization. Comprehensive experimental results show that the proposed method can indeed mitigate the overfitting problem in FSCIL, and outperform the state-of-the-art methods.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (59)
  1. Embracing change: Continual learning in deep neural networks. Trends in cognitive sciences, 24(12):1028–1040, 2020.
  2. A survey on transfer learning. IEEE Transactions on knowledge and data engineering, 22(10):1345–1359, 2009.
  3. Continual lifelong learning with neural networks: A review. Neural Networks, 113:54–71, 2019.
  4. Learning without forgetting. IEEE transactions on pattern analysis and machine intelligence, 40(12):2935–2947, 2017.
  5. Fearnet: Brain-inspired model for incremental learning. arXiv preprint arXiv:1711.10563, 2017.
  6. Cognitively-inspired model for incremental learning using a few examples. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, pages 222–223, 2020.
  7. Few-shot class-incremental learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 12183–12192, 2020.
  8. Mgsvf: Multi-grained slow vs. fast framework for few-shot class-incremental learning. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021.
  9. Anatomy of catastrophic forgetting: Hidden representations and task semantics. arXiv preprint arXiv:2007.07400, 2020.
  10. New insights on reducing abrupt representation change in online continual learning. arXiv preprint arXiv:2203.03798, 2022.
  11. icarl: Incremental classifier and representation learning. In Proceedings of the IEEE conference on Computer Vision and Pattern Recognition, pages 2001–2010, 2017.
  12. Learning a unified classifier incrementally via rebalancing. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 831–839, 2019.
  13. Incremental few-shot learning via vector quantization in deep embedded space. In International Conference on Learning Representations, 2020.
  14. Learnable expansion-and-compression network for few-shot class-incremental learning. arXiv preprint arXiv:2104.02281, 2021a.
  15. Few-shot lifelong learning. arXiv preprint arXiv:2103.00991, 2021.
  16. Subspace regularizers for few-shot class incremental learning. arXiv preprint arXiv:2110.07059, 2021.
  17. Overcoming catastrophic forgetting in incremental few-shot learning by finding flat minima. Advances in Neural Information Processing Systems, 34, 2021.
  18. Few-shot class-incremental learning via relation knowledge distillation. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 35, pages 1255–1263, 2021.
  19. Self-promoted prototype refinement for few-shot class-incremental learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 6801–6810, 2021.
  20. Learning adaptive classifiers synthesis for generalized few-shot learning. International Journal of Computer Vision, 129(6):1930–1953, 2021.
  21. Few-shot incremental learning with continually evolved classifiers. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 12455–12464, 2021.
  22. Thomas G Dietterich. Ensemble methods in machine learning. In International workshop on multiple classifier systems, pages 1–15. Springer, 2000.
  23. Few-shot image classification: Just use a library of pre-trained feature extractors and a simple classifier. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 9445–9454, 2021.
  24. Generalizing from a few examples: A survey on few-shot learning. ACM computing surveys (csur), 53(3):1–34, 2020.
  25. Prototypical networks for few-shot learning. Advances in neural information processing systems, 30, 2017.
  26. Learning to compare: Relation network for few-shot learning. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 1199–1208, 2018.
  27. Matching networks for one shot learning. Advances in neural information processing systems, 29, 2016.
  28. Deepemd: Few-shot image classification with differentiable earth mover’s distance and structured classifiers. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 12203–12213, 2020.
  29. Model-agnostic meta-learning for fast adaptation of deep networks. In International conference on machine learning, pages 1126–1135. PMLR, 2017.
  30. Task agnostic meta-learning for few-shot learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 11719–11727, 2019.
  31. A new meta-baseline for few-shot learning. 2020.
  32. Generalized and incremental few-shot learning by explicit learning and calibration without forgetting. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 9020–9029, 2021.
  33. A continual learning survey: Defying forgetting in classification tasks. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021.
  34. Class-incremental learning: survey and performance evaluation on image classification. arXiv preprint arXiv:2010.15277, 2020.
  35. Gradient episodic memory for continual learning. Advances in neural information processing systems, 30, 2017.
  36. Overcoming catastrophic forgetting for continual learning via model adaptation. In International conference on learning representations, 2018.
  37. Continual learning with deep generative replay. Advances in neural information processing systems, 30, 2017.
  38. Continual classification learning using generative models. arXiv preprint arXiv:1810.10612, 2018.
  39. Riemannian walk for incremental learning: Understanding forgetting and intransigence. In Proceedings of the European Conference on Computer Vision (ECCV), pages 532–547, 2018.
  40. Overcoming catastrophic forgetting in neural networks. Proceedings of the national academy of sciences, 114(13):3521–3526, 2017.
  41. End-to-end incremental learning. In Proceedings of the European conference on computer vision (ECCV), pages 233–248, 2018.
  42. Forward compatible few-shot class-incremental learning. arXiv preprint arXiv:2203.06953, 2022.
  43. Deep representation learning on long-tailed data: A learnable embedding augmentation perspective. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 2970–2979, 2020.
  44. Feature space augmentation for long-tailed data. In European Conference on Computer Vision, pages 694–710. Springer, 2020.
  45. Delta-encoder: an effective sample synthesis method for few-shot object recognition. Advances in Neural Information Processing Systems, 31, 2018.
  46. Metagan: An adversarial approach to few-shot learning. Advances in neural information processing systems, 31, 2018.
  47. Free lunch for few-shot learning: Distribution calibration. arXiv preprint arXiv:2101.06395, 2021b.
  48. Low-shot learning from imaginary data. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 7278–7286, 2018.
  49. Improved regularization of convolutional neural networks with cutout. arXiv preprint arXiv:1708.04552, 2017.
  50. mixup: Beyond empirical risk minimization. arXiv preprint arXiv:1710.09412, 2017.
  51. Cutmix: Regularization strategy to train strong classifiers with localizable features. In Proceedings of the IEEE/CVF international conference on computer vision, pages 6023–6032, 2019.
  52. Mixmo: Mixing multiple inputs for multiple outputs via deep subnetworks. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 823–833, 2021.
  53. Training independent subnetworks for robust prediction. arXiv preprint arXiv:2010.06610, 2020.
  54. Vicreg: Variance-invariance-covariance regularization for self-supervised learning. arXiv preprint arXiv:2105.04906, 2021.
  55. Learning multiple layers of features from tiny images. 2009.
  56. Imagenet large scale visual recognition challenge. International journal of computer vision, 115(3):211–252, 2015.
  57. The caltech-ucsd birds-200-2011 dataset. 2011.
  58. Semantic-aware knowledge distillation for few-shot class-incremental learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 2534–2543, 2021.
  59. Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 770–778, 2016.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (5)
  1. Mingli Zhu (12 papers)
  2. Zihao Zhu (39 papers)
  3. Sihong Chen (14 papers)
  4. Chen Chen (753 papers)
  5. Baoyuan Wu (107 papers)
Citations (2)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now