Hubs and Hyperspheres: Reducing Hubness and Improving Transductive Few-Shot Learning with Hyperspherical Embeddings
Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review
Standard
Hubs and Hyperspheres : Reducing Hubness and Improving Transductive Few-Shot Learning with Hyperspherical Embeddings. / Trosten, Daniel J.; Chakraborty, Rwiddhi; Loksc, Sigurd; Wickstrøm, Kristoffer Knutsen; Jenssen, Robert; Kampffmeyer, Michael C.
Proceedings - 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2023. IEEE Computer Society Press, 2023. p. 7527-7536.Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - GEN
T1 - Hubs and Hyperspheres
T2 - 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2023
AU - Trosten, Daniel J.
AU - Chakraborty, Rwiddhi
AU - Loksc, Sigurd
AU - Wickstrøm, Kristoffer Knutsen
AU - Jenssen, Robert
AU - Kampffmeyer, Michael C.
N1 - Publisher Copyright: © 2023 IEEE.
PY - 2023
Y1 - 2023
N2 - 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 11Code available at https://github.com/uitml/noHub..
AB - 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 11Code available at https://github.com/uitml/noHub..
KW - continual
KW - low-shot
KW - meta
KW - or long-tail learning
KW - Transfer
UR - http://www.scopus.com/inward/record.url?scp=85171965171&partnerID=8YFLogxK
U2 - 10.1109/CVPR52729.2023.00727
DO - 10.1109/CVPR52729.2023.00727
M3 - Article in proceedings
AN - SCOPUS:85171965171
SP - 7527
EP - 7536
BT - Proceedings - 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2023
PB - IEEE Computer Society Press
Y2 - 18 June 2023 through 22 June 2023
ER -
ID: 371288564