Semantics-Preserving Locality Embedding for Zero-Shot Learning

Shih-Yen Tao, Yi-Ren Yeh and Yu-Chiang Frank Wang

Abstract

Zero-shot learning (ZSL) aims at recognizing data as an unseen category, using information learned from the training data of predefined (seen) labels or attributes. In this paper, we propose an effective learning model for solving ZSL, which focuses on relating image and semantic domains with classification guarantees. In particular, we introduce semantics-preserving locality embedding when associating the above cross-domain data. We show that our ZSL model can be extended from inductive and transductive ZSL settings, if unlabeled data of unseen categories are presented during training.

Session

Spotlights

Files

PDF iconPaper (PDF)
PDF iconSupplementary (PDF)
MP4 iconVideo (MP4)

DOI

10.5244/C.31.3
https://dx.doi.org/10.5244/C.31.3

Citation

Shih-Yen Tao, Yi-Ren Yeh and Yu-Chiang Frank Wang. Semantics-Preserving Locality Embedding for Zero-Shot Learning. In T.K. Kim, S. Zafeiriou, G. Brostow and K. Mikolajczyk, editors, Proceedings of the British Machine Vision Conference (BMVC), pages 3.1-3.12. BMVA Press, September 2017.

Bibtex

            @inproceedings{BMVC2017_3,
                title={Semantics-Preserving Locality Embedding for Zero-Shot Learning},
                author={Shih-Yen Tao, Yi-Ren Yeh and Yu-Chiang Frank Wang},
                year={2017},
                month={September},
                pages={3.1-3.12},
                articleno={3},
                numpages={12},
                booktitle={Proceedings of the British Machine Vision Conference (BMVC)},
                publisher={BMVA Press},
                editor={Tae-Kyun Kim, Stefanos Zafeiriou, Gabriel Brostow and Krystian Mikolajczyk},
                doi={10.5244/C.31.3},
                isbn={1-901725-60-X},
                url={https://dx.doi.org/10.5244/C.31.3}
            }