Abstract
Generalized Few-Shot Learning (GFSL) applies the model trained with the base classes to predict the samples from both base classes and novel classes, where each novel class is only provided with a few labeled samples during testing. Limited by the severe data imbalance between base and novel classes, GFSL easily suffers from the prediction shift issue that most test samples tend to be classified into the base classes. Unlike the existing works that address this issue by either multi-stage training or complicated model design, we argue that extracting both discriminative and generalized feature representations is all GFSL needs, which could be achieved by simply scattering the intra-class distribution during training. Specifically, we introduce two self-supervised auxiliary tasks and a label permutation task to encourage the model to learn more image-level feature representations and push the decision boundary from novel towards base classes during inference. Our method is one-stage and could perform online inference. Experiments on the miniImageNet and tieredImageNet datasets show that the proposed method achieves comparable performance with the state-of-the-art multi-stage competitors under both traditional FSL and GFSL tasks, empirically proving that feature representation is the key for GFSL.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Ahn, S., Ko, J., Yun, S.Y.: Cuda: curriculum of data augmentation for long-tailed recognition. arXiv preprint arXiv:2302.05499 (2023)
Baik, S., Choi, J., Kim, H., Cho, D., Min, J., Lee, K.M.: Meta-learning with task-adaptive loss function for few-shot learning. In: ICCV, pp. 9465–9474 (2021)
Brinkmeyer, L., Drumond, R.R., Burchert, J., Schmidt-Thieme, L.: Few-shot forecasting of time-series with heterogeneous channels. In: ECML, pp. 3–18 (2023)
Cao, K., Wei, C., Gaidon, A., Arechiga, N., Ma, T.: Learning imbalanced datasets with label-distribution-aware margin loss. In: NeurIPS (2019)
Chawla, N.V., Bowyer, K.W., Hall, L.O., Kegelmeyer, W.P.: Smote: synthetic minority over-sampling technique. J. Artifi. Intell. Res. 16, 321–357 (2002)
Chen, Y., Liu, Z., Xu, H., Darrell, T., Wang, X.: Meta-baseline: exploring simple meta-learning for few-shot learning. In: ICCV, pp. 9062–9071 (2021)
Cheng, L., Fang, C., Zhang, D., Li, G., Huang, G.: Compound batch normalization for long-tailed image classification. In: ACM, pp. 1925–1934 (2022)
Cui, Y., Jia, M., Lin, T.Y., Song, Y., Belongie, S.: Class-balanced loss based on effective number of samples. In: CVPR, pp. 9268–9277 (2019)
Deng, J., Dong, W., Socher, R., Li, L.J., Li, K., Fei-Fei, L.: Imagenet: A large-scale hierarchical image database. In: CVPR, pp. 248–255 (2009)
Drummond, C., Holte, R.C., et al.: C4. 5, class imbalance, and cost sensitivity: why under-sampling beats over-sampling. In: Workshop on lEarning from Imbalanced Datasets II, pp. 1–8 (2003)
Gidaris, S., Bursuc, A., Komodakis, N., Pérez, P., Cord, M.: Boosting few-shot visual learning with self-supervision. In: ICCV, pp. 8059–8068 (2019)
Gidaris, S., Komodakis, N.: Dynamic few-shot visual learning without forgetting. In: CVPR, pp. 4367–4375 (2018)
Hariharan, B., Girshick, R.: Low-shot visual recognition by shrinking and hallucinating features. In: ICCV, pp. 3018–3027 (2017)
He, Y., et al.: Attribute surrogates learning and spectral tokens pooling in transformers for few-shot learning. In: CVPR, pp. 9119–9129 (2022)
Hinton, G., Vinyals, O., Dean, J.: Distilling the knowledge in a neural network. arXiv preprint arXiv:1503.02531 (2015)
Jian, Y., Torresani, L.: Label hallucination for few-shot classification. In: AAAI, pp. 7005–7014 (2022)
Kang, D., Kwon, H., Min, J., Cho, M.: Relational embedding for few-shot classification. In: ICCV, pp. 8822–8833 (2021)
Kukleva, A., Kuehne, H., Schiele, B.: Generalized and incremental few-shot learning by explicit learning and calibration without forgetting. In: ICCV, pp. 9020–9029 (2021)
Le, D., Nguyen, K.D., Nguyen, K., Tran, Q.H., Nguyen, R., Hua, B.S.: Poodle: Improving few-shot learning via penalizing out-of-distribution samples. In: NeurIPS, pp. 23942–23955 (2021)
Liu, J., Sun, Y., Han, C., Dou, Z., Li, W.: Deep representation learning on long-tailed data: A learnable embedding augmentation perspective. In: CVPR, pp. 2970–2979 (2020)
Ouali, Y., Hudelot, C., Tami, M.: Spatial contrastive learning for few-shot classification. In: ECML, pp. 671–686 (2021)
Park, S., Lim, J., Jeon, Y., Choi, J.Y.: Influence-balanced loss for imbalanced visual classification. In: ICCV, pp. 735–744 (2021)
Qi, H., Brown, M., Lowe, D.G.: Low-shot learning with imprinted weights. In: CVPR, pp. 5822–5830 (2018)
Rajasegaran, J., Khan, S., Hayat, M., Khan, F.S., Shah, M.: Self-supervised knowledge distillation for few-shot learning. arXiv preprint arXiv:2006.09785 (2020)
Rangwani, H., Aithal, S.K., Mishra, M., et al.: Escaping saddle points for effective generalization on class-imbalanced data. In: NeurIPS, pp. 22791–22805 (2022)
Ren, M., et al.: Meta-learning for semi-supervised few-shot classification. In: ICLR (2018)
Rizve, M.N., Khan, S., Khan, F.S., Shah, M.: Exploring complementary strengths of invariant and equivariant representations for few-shot learning. In: CVPR, pp. 10836–10846 (2021)
Selvaraju, R.R., Cogswell, M., Das, A., Vedantam, R., Parikh, D., Batra, D.: Grad-cam: Visual explanations from deep networks via gradient-based localization. In: ICCV, pp. 618–626 (2017)
Snell, J., Swersky, K., Zemel, R.: Prototypical networks for few-shot learning. In: NeurIPS (2017)
Su, J.C., Maji, S., Hariharan, B.: Boosting supervision with self-supervision for few-shot learning. In: ICCV, pp. 8059–8068 (2019)
Su, J.-C., Maji, S., Hariharan, B.: When does self-supervision improve few-shot learning? In: Vedaldi, A., Bischof, H., Brox, T., Frahm, J.-M. (eds.) ECCV 2020. LNCS, vol. 12352, pp. 645–666. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-58571-6_38
Tian, Y., Wang, Y., Krishnan, D., Tenenbaum, J.B., Isola, P.: Rethinking few-shot image classification: a good embedding is all you need? In: Vedaldi, A., Bischof, H., Brox, T., Frahm, J.-M. (eds.) ECCV 2020. LNCS, vol. 12359, pp. 266–282. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-58568-6_16
Vinyals, O., Blundell, C., Lillicrap, T., Wierstra, D., et al.: Matching networks for one-shot learning. In: NeurIPS, pp. 3630–3638 (2016)
Xie, J., Long, F., Lv, J., Wang, Q., Li, P.: Joint distribution matters: Deep brownian distance covariance for few-shot classification. In: CVPR, pp. 7972–7981 (2022)
Ye, H.J., Hu, H., Zhan, D.C.: Learning adaptive classifiers synthesis for generalized few-shot learning. IJCV 129, 1930–1953 (2021)
Ye, H.J., Hu, H., Zhan, D.C., Sha, F.: Few-shot learning via embedding adaptation with set-to-set functions. In: CVPR, pp. 8808–8817 (2020)
Yu, T., He, S., Song, Y.Z., Xiang, T.: Hybrid graph neural networks for few-shot learning. In: AAAI, pp. 3179–3187 (2022)
Yu, Y., Zhang, D., Ji, Z.: Masked feature generation network for few-shot learning
Zhang, T., Huang, W.: Kernel relative-prototype spectral filtering for few-shot learning. In: ECCV, pp. 541–557 (2022). https://doi.org/10.1007/978-3-031-20044-1_31
Acknowledgments
This work is supported in part by the Key R &D Program of Zhejiang Province, China (2023C01043, 2021C01119) and the National Natural Science Foundation of China under Grant (62002320, U19B2043, 61672456).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Ethics declarations
Ethical Statement
In this paper, we mainly investigate the problem of how to balance the base classes and novel classes in the generalized FSL task and propose that what the GFSL needs is extracting both discriminative and generalized feature representations. We consider that the required discussion about ethic and future societal effect is not applicable for our work.
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Yu, Y., Jin, L., Li, Y. (2023). Boosting Generalized Few-Shot Learning by Scattering Intra-class Distribution. In: Koutra, D., Plant, C., Gomez Rodriguez, M., Baralis, E., Bonchi, F. (eds) Machine Learning and Knowledge Discovery in Databases: Research Track. ECML PKDD 2023. Lecture Notes in Computer Science(), vol 14170. Springer, Cham. https://doi.org/10.1007/978-3-031-43415-0_26
Download citation
DOI: https://doi.org/10.1007/978-3-031-43415-0_26
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-43414-3
Online ISBN: 978-3-031-43415-0
eBook Packages: Computer ScienceComputer Science (R0)