Federated Learning and Optimization for Few-Shot Image Classification
Source abstract
Abstract
Image classification is crucial for various applications, including digital construction, smart manufacturing, and medical imaging. Focusing on the inadequate model generalization and data privacy concerns in few-shot image classification, in this paper, we propose a federated learning approach that incorporates privacy-preserving techniques. First, we utilize contrastive learning to train on local few-shot image data and apply various data augmentation methods to expand the sample size, thereby enhancing the model’s generalization capabilities in few-shot contexts. Second, we introduce local differential privacy techniques and weight pruning methods to safeguard model parameters, perturbing the transmitted parameters to ensure user data privacy. Finally, numerical simulations are conducted to demonstrate the effectiveness of our proposed method. The results indicate that our approach significantly enhances model generalization and test accuracy compared to several popular federated learning algorithms while maintaining data privacy, highlighting its effectiveness and practicality in addressing the challenges of model generalization and data privacy in few-shot image scenarios.
Source: openalex_inverted_index · Quality: baseline
Research context
OpenAlex topics: Privacy-Preserving Technologies in Data · Brain Tumor Detection and Classification · Stochastic Gradient Optimization Techniques
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (49)
Show source-provided references
1. Efficient hybrid oversampling and intelligent undersampling for imbalanced big data classification10.1016/j.eswa.2024.123149 ↗
2. Learning from massive noisy labeled data for image classification
3. Dhillon GS, Chaudhari P, Ravichandran A, Soatto S. A baseline for few-shot image classification. 2019. doi:10.48550/arXiv.1909.02729.
4. Contrastive learning of medical visual representations from paired images and text
5. Structure-aware prototypical neural process for few-shot graph classification10.1109/TNNLS.2022.3173318 ↗
6. Rethinking few-shot image classification: a good embedding is all you need?
7. A review of applications in federated learning10.1016/j.cie.2020.106854 ↗
8. Communication-efficient learning of deep networks from decentralized data
9. A comprehensive survey of few-shot learning: evolution, applications, challenges, and opportunities10.1145/3582688 ↗
10. A simple framework for contrastive learning of visual representations
11. Comprehensive privacy analysis of deep learning
12. FL-EASGD: federated learning privacy security method based on homomorphic encryption
13. Federated optimization in heterogeneous networks
14. Differentially private meta-learning
15. Federated few-shot learning with adversarial learning
16. Federated few-shot learning
17. Finding trustworthy neighbors: graph aided federated learning for few- shot industrial fault diagnosis with data heterogeneity10.1016/j.jprocont.2023.103038 ↗
18. Federated few-shot learning for cough classification with edge devices10.1007/s10489-023-05006-4 ↗
19. Personalized federated few-shot learning10.1109/TNNLS.2022.3190359 ↗
20. AugMix: a simple data processing method to improve robustness and uncertainty
21. DeVries T, Taylor GW. Improved regularization of convolutional neural networks with cutout. 2017. doi:10.48550/arXiv.1708.04552.
22. CutMix: regularization strategy to train strong classifiers with localizable features
23. MaskMix: mask mixing augmentation method for change detection
24. PatchMix augmentation to identify causal features in few-shot learning
25. Differential privacy
26. Differential privacy techniques for cyber physical systems: a survey10.1109/COMST.2019.2944748 ↗
27. Collecting and analyzing multidimensional data with local differential privacy
28. LDP-Feat: image features with local differential privacy
29. A LDP-based privacy-preserving longitudinal and multidimensional range query scheme in IOT10.1109/JIOT.2023.3306003 ↗
30. LDP-FL: practical private aggregation in federated learning with local differential privacy
31. What makes for good views for contrastive learning?
32. Fair contrastive learning for facial attribute classification
33. Mechanism design via differential privacy
34. TADAM: task dependent adaptive metric for improved few- shot learning
35. Chen X, Fan H, Girshick R, He K. Improved baselines with momentum contrastive learning. 2020. doi:10.48550/arXiv.2003.04297.
36. Learning a few-shot embedding model with contrastive learning
37. Matching networks for one shot learning
38. ImageNet large scale visual recognition challenge10.1007/s11263-015-0816-y ↗
39. Model-agnostic meta-learning for fast adaptation of deep networks
40. Generative adversarial networks10.1145/3422622 ↗
41. Large-scale machine learning with stochastic gradient descent
42. Goyal P, Dollár P, Girshick R, Noordhuis P, Wesolowski L, Kyrola A, et al. Large minibatch SGD: training imageNet in 1 hour. 2010. doi:10.48550/arXiv.1706.02677.
43. Prototypical networks for few-shot learning
44. Hsu TH, Qi H, Brown M. Measuring the effects of non-identical data distribution for federated visual classification. 2019. doi:10.48550/arXiv.1909.06335.
45. Yu T, Bagdasaryan E, Shmatikov V. Salvaging federated learning by local adaptation. 2020. doi:10.48550/arXiv.2002.04758.
46. Deep learning with differential privacy
47. McMahan HB, Andrew G, Erlingsson U, Chien S, Mironov I, Papernot N, et al. A general approach to adding differential privacy to iterative training procedures. 2018. doi:10.48550/arXiv.1812.06210.
48. Local differential privacy-based federated learning for internet of things10.1109/JIOT.2020.3037194 ↗
49. DP-FedIOD: a differential privacy and federated learning based framework for aerial insulators orientation detection10.1109/ACCESS.2024.3380195 ↗