Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
80 tokens/sec
GPT-4o
59 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
7 tokens/sec
GPT-4.1 Pro
50 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

GCT: Graph Co-Training for Semi-Supervised Few-Shot Learning (2203.07738v4)

Published 15 Mar 2022 in cs.CV

Abstract: Few-shot learning (FSL), purposing to resolve the problem of data-scarce, has attracted considerable attention in recent years. A popular FSL framework contains two phases: (i) the pre-train phase employs the base data to train a CNN-based feature extractor. (ii) the meta-test phase applies the frozen feature extractor to novel data (novel data has different categories from base data) and designs a classifier for recognition. To correct few-shot data distribution, researchers propose Semi-Supervised Few-Shot Learning (SSFSL) by introducing unlabeled data. Although SSFSL has been proved to achieve outstanding performances in the FSL community, there still exists a fundamental problem: the pre-trained feature extractor can not adapt to the novel data flawlessly due to the cross-category setting. Usually, large amounts of noises are introduced to the novel feature. We dub it as Feature-Extractor-Maladaptive (FEM) problem. To tackle FEM, we make two efforts in this paper. First, we propose a novel label prediction method, Isolated Graph Learning (IGL). IGL introduces the Laplacian operator to encode the raw data to graph space, which helps reduce the dependence on features when classifying, and then project graph representation to label space for prediction. The key point is that: IGL can weaken the negative influence of noise from the feature representation perspective, and is also flexible to independently complete training and testing procedures, which is suitable for SSFSL. Second, we propose Graph Co-Training (GCT) to tackle this challenge from a multi-modal fusion perspective by extending the proposed IGL to the co-training framework. GCT is a semi-supervised method that exploits the unlabeled samples with two modal features to crossly strengthen the IGL classifier.

PDF HTML Abstract
Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Bookmark Chat Bubble Oval Streamline Icon: https://streamlinehq.com Chat (Pro)
Definition Search Book Streamline Icon: https://streamlinehq.com
References (33)
  1. Laplacian eigenmaps and spectral techniques for embedding and clustering. In Neural Information Processing Systems, pages 585–591, 2002.
  2. Mixmatch: A holistic approach to semi-supervised learning. In Neural Information Processing Systems, pages 5049–5059, 2019.
  3. Metamix: Improved meta-learning with interpolation-based consistency regularization. In ICPR, 2020.
  4. Diversity with cooperation: Ensemble methods for few-shot classification. In International Conference on Computer Vision, pages 3723–3731, 2019.
  5. Selecting relevant features from a multi-domain representation for few-shot classification. In European Conference on Computer Vision, pages 769–786. Springer, 2020.
  6. Contextual multi-scale feature learning for person re-identification. In ACM International Conference on Multimedia, pages 655–663, 2020.
  7. Few-shot learning with graph neural networks. In International Conference on Learning Representations, 2017.
  8. Pseudo-loss confidence metric for semi-supervised few-shot learning. In International Conference on Computer Vision, pages 8671–8680, 2021.
  9. Behavior regularized prototypical networks for semi-supervised few-shot image classification. Pattern Recognition, page 107765, 2020.
  10. Edge-labeling graph neural network for few-shot learning. In Computer Vision and Pattern Recognition, pages 11–20, 2019.
  11. Iterative label cleaning for transductive and semi-supervised few-shot learning. In International Conference on Computer Vision, pages 8751–8760, 2021.
  12. Meta-learning with differentiable convex optimization. In Computer Vision and Pattern Recognition, pages 10657–10665, 2019.
  13. Learning to self-train for semi-supervised few-shot classification. In Neural Information Processing Systems, pages 10276–10286, 2019.
  14. Dense classification and implanting for few-shot learning. In Computer Vision and Pattern Recognition, pages 9258–9267, 2019.
  15. A universal representation transformer layer for few-shot image classification. In International Conference on Learning Representations, 2021.
  16. Charting the right manifold: Manifold mixup for few-shot learning. In IEEE Winter Conference on Applications of Computer Vision, pages 2218–2227, 2020.
  17. Assanet: An anisotropic separable set abstraction for efficient point cloud representation learning. In Advances in Neural Information Processing Systems, 2021.
  18. Label embedded dictionary learning for image classification. Neurocomputing, 385:122–131, 2020.
  19. Mhfc: Multi-head feature collaboration for few-shot learning. In ACM International Conference on Multimedia, 2021a.
  20. Mdfm: Multi-decision fusing model for few-shot learning. IEEE Transactions on Circuits and Systems for Video Technology, 2021b.
  21. Fixmatch: Simplifying semi-supervised learning with consistency and confidence. In Neural Information Processing Systems, 2020.
  22. Mutual crf-gnn for few-shot learning. In Conference on Computer Vision and Pattern Recognition, pages 2329–2339, 2021.
  23. Cost-effective active learning for deep image classification. IEEE Transactions on Circuits and Systems for Video Technology, 27(12):2591–2600, 2016.
  24. Dense-scale feature learning in person re-identification. In Asian Conference on Computer Vision, 2020a.
  25. Instance credibility inference for few-shot learning. In Computer Vision and Pattern Recognition, pages 12836–12845, 2020b.
  26. Rotation-invariant point cloud representation for 3-d model recognition. IEEE Transactions on Cybernetics, 2022.
  27. Linkage based face clustering via graph convolution network. In Conference on Computer Vision and Pattern Recognition, pages 1117–1125, 2019.
  28. Learning to cluster faces via confidence and connectivity estimation. In Conference on Computer Vision and Pattern Recognition, pages 13369–13378, 2020a.
  29. Dpgn: Distribution propagation graph network for few-shot learning. In Computer Vision and Pattern Recognition, pages 13390–13399, 2020b.
  30. Inductive multi-hypergraph learning and its application on view-based 3d object classification. IEEE Transactions on Image Processing, 27(12):5957–5968, 2018.
  31. Real-time scene-aware lidar point cloud compression using semantic prior representation. IEEE Transactions on Circuits and Systems for Video Technology, 2022.
  32. Pedestrian alignment network for large-scale person re-identification. IEEE Transactions on Circuits and Systems for Video Technology, 29(10):3037–3045, 2018.
  33. Learning with local and global consistency. Neural Information Processing Systems, 16:321–328, 2003.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (7)
  1. Rui Xu (199 papers)
  2. Lei Xing (83 papers)
  3. Shuai Shao (57 papers)
  4. Lifei Zhao (3 papers)
  5. Baodi Liu (4 papers)
  6. Weifeng Liu (46 papers)
  7. Yicong Zhou (26 papers)
Citations (20)
View on Semantic Scholar