Co-consistent regularization with discriminative feature for zero-shot learning

Yanling Tian, Weitong Zhang, Qieshi Zhang, Jun Cheng, Pengyi Hao, Gang Lu

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

With the development of deep learning, zero-shot learning (ZSL) issues deserve more attention. Due to the problems of projection domain shift and discriminative feature extraction, we propose an end-to-end framework, which is different from traditional ZSL methods in the following two aspects: (1) we use a cascaded network to automatically locate discriminative regions, which can better extract latent features and contribute to the representation of key semantic attributes. (2) our framework achieves mapping in visual-semantic embedding space and calculation procedure of the dot product in deep learning framework. In addition, a joint loss function is designed for the regularization constraint of the whole method and achieves supervised learning, which enhances generalization ability in test set. In this paper, we make some experiments on Animals with Attributes 2 (AwA2), Caltech-UCSD Birds 200-2011 (CUB) and SUN datasets, which achieves better results compared to the state-of-the-art methods.

Original languageEnglish
Title of host publicationNeural Information Processing - 25th International Conference, ICONIP 2018, Proceedings
EditorsLong Cheng, Andrew Chi Sing Leung, Seiichi Ozawa
PublisherSpringer Verlag
Pages33-45
Number of pages13
ISBN (Print)9783030041663
DOIs
Publication statusPublished - 2018
Externally publishedYes
Event25th International Conference on Neural Information Processing, ICONIP 2018 - Siem Reap, Cambodia
Duration: 2018 Dec 132018 Dec 16

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11301 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other25th International Conference on Neural Information Processing, ICONIP 2018
CountryCambodia
CitySiem Reap
Period18/12/1318/12/16

Keywords

  • Discriminative region
  • Projection domain shift
  • Regularization
  • Supervised learning
  • Zero-shot learning (ZSL)

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Co-consistent regularization with discriminative feature for zero-shot learning'. Together they form a unique fingerprint.

  • Cite this

    Tian, Y., Zhang, W., Zhang, Q., Cheng, J., Hao, P., & Lu, G. (2018). Co-consistent regularization with discriminative feature for zero-shot learning. In L. Cheng, A. C. S. Leung, & S. Ozawa (Eds.), Neural Information Processing - 25th International Conference, ICONIP 2018, Proceedings (pp. 33-45). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11301 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-04167-0_4