mlsanityData to 7 Oct 2026

Hubs and Hyperspheres: Reducing Hubness and Improving Transductive Few-shot Learning with Hyperspherical Embeddings

arXiv 2303.09352 ↗First posted 16 Mar 2023cs.CVcited by 11

Not on the map. Why?

Abstract (arXiv)

Distance-based classification is frequently used in transductive few-shot learning (FSL). However, due to the high-dimensionality of image representations, FSL classifiers are prone to suffer from the hubness problem, where a few points (hubs) occur frequently in multiple nearest neighbour lists of other points. Hubness negatively impacts distance-based classification when hubs from one class appear often among the nearest neighbors of points from another class, degrading the classifier's performance. To address the hubness problem in FSL, we first prove that hubness can be eliminated by distributing representations uniformly on the hypersphere. We then propose two new approaches to embed representations on the hypersphere, which we prove optimize a tradeoff between uniformity and local similarity preservation -- reducing hubness while retaining class structure. Our experiments show that the proposed methods reduce hubness, and significantly improves transductive FSL accuracy for a wide range of classifiers.

Key facts. Cited by 11 papers in the map (data through 7 Oct 2026). Builds on 26 of its 45 references that are in the map. First posted on arXiv 16 Mar 2023.

Builds on

26 of 45 references.

References · 26

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  3. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  4. 2016Wide Residual Networkscited by 3,057
  5. 2016Matching Networks for One Shot Learningcited by 2,246
  6. 2018Unsupervised Representation Learning by Predicting Image Rotationscited by 1,586
  7. 2020Understanding Contrastive Representation Learning through Alignment and Uniformity on the Hyperspherecited by 1,393
  8. 2018Meta-Learning for Semi-Supervised Few-Shot Classificationcited by 456
  9. 2018Few-Shot Learning via Embedding Adaptation with Set-to-Set Functionscited by 293
  10. 2019SimpleShot: Revisiting Nearest-Neighbor Classification for Few-Shot Learningcited by 194
  11. 2019Charting the Right Manifold: Manifold Mixup for Few-shot Learningcited by 135
  12. 2022Pushing the Limits of Simple Pipelines for Few-Shot Learning: External Data and Fine-Tuning Make a Differencecited by 123
14 more
  1. 2019Edge-labeling Graph Neural Network for Few-shot Learningcited by 114
  2. 2019Infinite Mixture Prototypes for Few-Shot Learningcited by 99
  3. 2020Laplacian Regularized Few-Shot Learningcited by 90
  4. 2015Ridge Regression, Hubness, and Zero-Shot Learningcited by 44
  5. 2020Transductive Information Maximization For Few-Shot Learningcited by 44
  6. 2020Iterative label cleaning for transductive and semi-supervised few-shot learningcited by 40
  7. 2022Realistic Evaluation of Transductive Few-Shot Learningcited by 26
  8. 2021Transductive Few-Shot Classification on the Oblique Manifoldcited by 23
  9. 2022Alleviating the Sample Selection Bias in Few-shot Learning by Removing Projection to the Centroidcited by 20
  10. 2022POODLE: Improving Few-shot Learning via Penalizing Out-of-Distribution Samplescited by 15
  11. 2021The Role of Global Labels in Few-Shot Classification and How to Infer Themcited by 12
  12. 2022CAD: Co-Adapting Discriminative Features for Improved Few-Shot Classificationcited by 9
  13. 2022Powering Finetuning in Few-Shot Learning: Domain-Agnostic Bias Reduction with Selected Samplingcited by 8
  14. 2015VMF-SNE: Embedding for Spherical Datacited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 2; 2024: 1; 2025: 6; 2026: 2.

11 papers.

  1. 2024Transductive Zero-Shot and Few-Shot CLIPcited by 31
  2. 2023Contrastive Tuning: A Little Help to Make Masked Autoencoders Forgetcited by 15
  3. 2025Position: Beyond Euclidean -- Foundation Models Should Embrace Non-Euclidean Geometriescited by 15
  4. 2025Improving Open-world Continual Learning under the Constraints of Scarce Labeled Datacited by 2
  5. 2025Keep your distance: learning dispersed embeddings on mathbbSmcited by 1
  6. 2026Polaris: Coupled Orbital Polar Embeddings for Hierarchical Concept Learningcited by 1
  7. 2023Robust Transductive Few-shot Learning via Joint Message Passing and Prototype-based Soft-label Propagationcited by 0
  8. 2025Unbiased Max-Min Embedding Classification for Transductive Few-Shot Learning: Clustering and Classification Are All You Needcited by 0
  9. 2025HCFSLN: Adaptive Hyperbolic Few-Shot Learning for Multimodal Anxiety Detectioncited by 0
  10. 2025Query-aware Hub Prototype Learning for Few-Shot 3D Point Cloud Semantic Segmentationcited by 0
  11. 2026Torus embeddingscited by 0