mlsanityData to 7 Oct 2026

Not All Instances Contribute Equally: Instance-adaptive Class Representation Learning for Few-Shot Visual Recognition

arXiv 2209.03034 ↗First posted 7 Sep 2022cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Few-shot visual recognition refers to recognize novel visual concepts from a few labeled instances. Many few-shot visual recognition methods adopt the metric-based meta-learning paradigm by comparing the query representation with class representations to predict the category of query instance. However, current metric-based methods generally treat all instances equally and consequently often obtain biased class representation, considering not all instances are equally significant when summarizing the instance-level representations for the class-level representation. For example, some instances may contain unrepresentative information, such as too much background and information of unrelated concepts, which skew the results. To address the above issues, we propose a novel metric-based meta-learning framework termed instance-adaptive class representation learning network (ICRL-Net) for few-shot visual recognition. Specifically, we develop an adaptive instance revaluing network with the capability to address the biased representation issue when generating the class representation, by learning and assigning adaptive weights for different instances according to their relative significance in the support set of corresponding class. Additionally, we design an improved bilinear instance representation and incorporate two novel structural losses, i.e., intra-class instance clustering loss and inter-class representation distinguishing loss, to further regulate the instance revaluation process and refine the class representation. We conduct extensive experiments on four commonly adopted few-shot benchmarks: miniImageNet, tieredImageNet, CIFAR-FS, and FC100 datasets. The experimental results compared with the state-of-the-art approaches demonstrate the superiority of our ICRL-Net.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 49 of its 80 references that are in the map. First posted on arXiv 7 Sep 2022.

Builds on

49 of 80 references.

References · 49

  1. 2017Attention Is All You Needcited by 59,932
  2. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  3. 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,617
  4. 2017Graph Attention Networkscited by 8,854
  5. 2014ImageNet Large Scale Visual Recognition Challengecited by 8,441
  6. 2017Mask R-CNNcited by 7,259
  7. 2016Densely Connected Convolutional Networkscited by 6,691
  8. 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,861
  9. 2015You Only Look Once: Unified, Real-Time Object Detectioncited by 4,829
  10. 2017Squeeze-and-Excitation Networkscited by 4,763
  11. 2017Prototypical Networks for Few-shot Learningcited by 3,001
  12. 2018CBAM: Convolutional Block Attention Modulecited by 2,463
37 more
  1. 2017Non-local Neural Networkscited by 2,290
  2. 2016Matching Networks for One Shot Learningcited by 2,246
  3. 2017Learning to Compare: Relation Network for Few-Shot Learningcited by 1,139
  4. 2019A Closer Look at Few-shot Classificationcited by 679
  5. 2021Transformer in Transformercited by 632
  6. 2013Network In Networkcited by 600
  7. 2018Meta-Learning with Latent Embedding Optimizationcited by 476
  8. 2019Meta-Learning with Differentiable Convex Optimizationcited by 461
  9. 2018Meta-Learning for Semi-Supervised Few-Shot Classificationcited by 456
  10. 2020Rethinking Few-Shot Image Classification: a Good Embedding Is All You Need?cited by 435
  11. 2018TADAM: Task dependent adaptive metric for improved few-shot learningcited by 408
  12. 2018Meta-learning with differentiable closed-form solverscited by 395
  13. 2018Dynamic Few-Shot Visual Learning without Forgettingcited by 385
  14. 2017Meta-SGD: Learning to Learn Quickly for Few-Shot Learningcited by 377
  15. 2018Meta-Transfer Learning for Few-Shot Learningcited by 306
  16. 2018Few-Shot Learning via Embedding Adaptation with Set-to-Set Functionscited by 293
  17. 2015Bilinear CNNs for Fine-grained Visual Recognitioncited by 228
  18. 2019Cross Attention Network for Few-shot Classificationcited by 218
  19. 2020Meta-Baseline: Exploring Simple Meta-Learning for Few-Shot Learningcited by 179
  20. 2019Revisiting Local Descriptor based Image-to-Class Measure for Few-shot Learningcited by 121
  21. 2020Adversarial Feature Hallucination Networks for Few-Shot Learningcited by 110
  22. 2020Laplacian Regularized Few-Shot Learningcited by 90
  23. 2019Meta-Learning of Neural Architectures for Few-Shot Learningcited by 73
  24. 2019Generating Classification Weights with GNN Denoising Autoencoders for Few-Shot Learningcited by 69
  25. 2020Instance Credibility Inference for Few-Shot Learningcited by 57
  26. 2021Pareto Self-Supervised Training for Few-Shot Learningcited by 52
  27. 2019Few-Shot Learning via Saliency-guided Hallucination of Samplescited by 49
  28. 2020Unraveling Meta-Learning: Understanding Feature Representations for Few-Shot Taskscited by 47
  29. 2019Few-Shot Learning with Localization in Realistic Settingscited by 34
  30. 2020Power Normalizations in Fine-grained Image, Few-shot Image and Graph Classificationcited by 27
  31. 2019A Two-Stage Approach to Few-Shot Learning for Image Recognitioncited by 18
  32. 2020How to trust unlabeled data? Instance Credibility Inference for Few-Shot Learningcited by 16
  33. 2018Power Normalizing Second-order Similarity Network for Few-shot Learningcited by 13
  34. 2020Few-Shot Learning with Geometric Constraintscited by 13
  35. 2020Uncertainty-Aware Few-Shot Image Classificationcited by 9
  36. 2020ReMP: Rectified Metric Propagation for Few-Shot Learningcited by 2
  37. 2021Multi-Pretext Attention Network for Few-shot Learning with Self-supervisioncited by 2

Built on by

Not cited yet.