CLAREL: Classification via Retrieval Loss for Zero-Shot Learning

Boris N. Oreshkin, Negar Rostamzadeh, Pedro O. Pinheiro, Christopher Pal; Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2020, pp. 916-917

Abstract


We address the problem of learning cross-modal representations. We propose an instance-based deep metric learning approach in joint visual and textual space. The key novelty of this paper is that it shows that using per-image semantic supervision leads to substantial improvement in zero-shot performance over using class-only supervision. We also provide a probabilistic justification and empirical validation for a metric rescaling approach to balance the seen/unseen accuracy in the GZSL task. We evaluate our approach on two fine-grained zero-shot datasets: CUB and FLOWERS.

Related Material


[pdf] [supp]
[bibtex]
@InProceedings{Oreshkin_2020_CVPR_Workshops,
author = {Oreshkin, Boris N. and Rostamzadeh, Negar and Pinheiro, Pedro O. and Pal, Christopher},
title = {CLAREL: Classification via Retrieval Loss for Zero-Shot Learning},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops},
month = {June},
year = {2020}
}