From Classical to Generalized Zero-Shot Learning: A Simple Adaptation Process

Abstract : Zero-shot learning (ZSL) is concerned with the recognition of previously unseen classes. It relies on additional semantic knowledge for which a mapping can be learned with training examples of seen classes. While classical ZSL considers the recognition performance on unseen classes only, generalized zero-shot learning (GZSL) aims at maximizing performance on both seen and unseen classes. In this paper, we propose a new process for training and evaluation in the GZSL setting; this process addresses the gap in performance between samples from unseen and seen classes by penalizing the latter, and enables to select hyper-parameters well-suited to the GZSL task. It can be applied to any existing ZSL approach and leads to a significant performance boost: the experimental evaluation shows that GZSL performance, averaged over eight state-of-the-art methods, is improved from 28.5 to 42.2 on CUB and from 28.2 to 57.1 on AwA2.
Complete list of metadatas

Cited literature [31 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01983612
Contributor : Herve Le Borgne <>
Submitted on : Wednesday, January 16, 2019 - 3:57:23 PM
Last modification on : Monday, February 10, 2020 - 6:12:35 PM

File

lecacheux19arxiv.pdf
Files produced by the author(s)

Identifiers

Citation

Yannick Le Cacheux, Hervé Le Borgne, Michel Crucianu. From Classical to Generalized Zero-Shot Learning: A Simple Adaptation Process. Ioannis Kompatsiaris, Benoit Huet, Vasileios Mezaris, Cathal Gurrin, Wen-Huang Cheng, Stefanos Vrochidis. MultiMedia Modeling. 25th International Conference, MMM 2019, Thessaloniki, Greece, January 8–11, 2019, Proceedings, Part II, 11296, Springer Verlag, pp.465-477, 2019, Lecture Notes in Computer Science, 978-3-030-05716-9. ⟨10.1007/978-3-030-05716-9_38⟩. ⟨hal-01983612⟩

Share

Metrics

Record views

124

Files downloads

164