Toggle Main Menu Toggle Search

Open Access padlockePrints

Variational Disentangle Zero-Shot Learning

Lookup NU author(s): Dr Jie Su

Downloads


Licence

This work is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0).


Abstract

© 2023 by the authors. Existing zero-shot learning (ZSL) methods typically focus on mapping from the feature space (e.g., visual space) to class-level attributes, often leading to a non-injective projection. Such a mapping may cause a significant loss of instance-level information. While an ideal projection to instance-level attributes would be desirable, it can also be prohibitively expensive and thus impractical in many scenarios. In this work, we propose a variational disentangle zero-shot learning (VDZSL) framework that addresses this problem by constructing variational instance-specific attributes from a class-specific semantic latent distribution. Specifically, our approach disentangles each instance into class-specific attributes and the corresponding variant features. Unlike transductive ZSL, which assumes that unseen classes’ attributions are known beforehand, our VDZSL method does not rely on this strong assumption, making it more applicable in real-world scenarios. Extensive experiments conducted on three popular ZSL benchmark datasets (i.e., AwA2, CUB, and FLO) validate the effectiveness of our approach. In the conventional ZSL setting, our method demonstrates an improvement of 12∼15% relative to the advanced approaches and achieves a classification accuracy of 70% on the AwA2 dataset. Furthermore, under the more challenging generalized ZSL setting, our approach can gain an improvement of 5∼15% compared with the advanced methods.


Publication metadata

Author(s): Su J, Wan J, Li T, Li X, Ye Y

Publication type: Article

Publication status: Published

Journal: Mathematics

Year: 2023

Volume: 11

Issue: 16

Online publication date: 18/08/2024

Acceptance date: 16/08/2024

Date deposited: 02/10/2024

ISSN (electronic): 2227-7390

Publisher: MDPI

URL: https://doi.org/10.3390/math11163578

DOI: 10.3390/math11163578

Data Access Statement: No new data were created or analyzed in this study. Data sharing is not applicable to this article.


Altmetrics

Altmetrics provided by Altmetric


Share