Feature Redirection Network for Few-Shot Classification
Publication
, Chapter
Wang, Y; Zhong, G; Mao, Y; Huang, K
January 1, 2020
Few-shot classification aims to learn novel categories by giving few labeled samples. How to make best use of the limited data to obtain a learner with fast learning ability has become a challenging problem. In this paper, we propose a feature redirection network (FRNet) for few-shot classification to make the features more discriminative. The proposed FRNet not only highlights relevant category features of support samples, but also learns how to generate task-relevant features of query samples. Experiments conducted on three datasets have demonstrate its superiority over the state-of-the-art methods.
Duke Scholars
DOI
ISBN
9783030638191
Publication Date
January 1, 2020
Volume
1332
Start / End Page
418 / 425
Citation
APA
Chicago
ICMJE
MLA
NLM
Wang, Y., Zhong, G., Mao, Y., & Huang, K. (2020). Feature Redirection Network for Few-Shot Classification (Vol. 1332, pp. 418–425). https://doi.org/10.1007/978-3-030-63820-7_48
Wang, Y., G. Zhong, Y. Mao, and K. Huang. “Feature Redirection Network for Few-Shot Classification,” 1332:418–25, 2020. https://doi.org/10.1007/978-3-030-63820-7_48.
Wang Y, Zhong G, Mao Y, Huang K. Feature Redirection Network for Few-Shot Classification. In 2020. p. 418–25.
Wang, Y., et al. Feature Redirection Network for Few-Shot Classification. Vol. 1332, 2020, pp. 418–25. Scopus, doi:10.1007/978-3-030-63820-7_48.
Wang Y, Zhong G, Mao Y, Huang K. Feature Redirection Network for Few-Shot Classification. 2020. p. 418–425.
DOI
ISBN
9783030638191
Publication Date
January 1, 2020
Volume
1332
Start / End Page
418 / 425