Diversity-Boosted Generalization-Specialization Balancing for Zero-Shot Learning

Li, Y; Liu, Z; Chang, X; McAuley, J; Yao, L

Diversity-Boosted Generalization-Specialization Balancing for Zero-Shot Learning

Li, Y Liu, Z Chang, X

McAuley, J Yao, L

Permalink

Publisher:: Institute of Electrical and Electronics Engineers (IEEE)
Publication Type:: Journal Article
Citation:: IEEE Transactions on Multimedia, 2023, PP, (99), pp. 1-11
Issue Date:: 2023-01-01

Embargoed

	Filename	Description	Size
	Diversity-Boosted Generalization.pdf	Submitted version	1 MB	Adobe PDF	View/Open

Copyright Clearance Process

Recently Added
In Progress
Embargoed
Open Access

This item is currently unavailable due to the publisher's embargo.

The embargo period expires on 11 Jan 2025

Full metadata record

Field	Value	Language
dc.contributor.author	Li, Y
dc.contributor.author	Liu, Z
dc.contributor.author	Chang, X https://orcid.org/0000-0002-7778-8807
dc.contributor.author	McAuley, J
dc.contributor.author	Yao, L
dc.date.accessioned	2023-03-31T09:41:06Z
dc.date.available	2023-03-31T09:41:06Z
dc.date.issued	2023-01-01
dc.identifier.citation	IEEE Transactions on Multimedia, 2023, PP, (99), pp. 1-11
dc.identifier.issn	1520-9210
dc.identifier.issn	1941-0077
dc.identifier.uri	http://hdl.handle.net/10453/168969
dc.description.abstract	Zero-Shot Learning (ZSL) aims to transfer classification capability from seen to unseen classes. Recent methods have proved that generalization and specialization are two essential abilities to achieve good performance in ZSL. However, focusing on only one of the abilities may result in models that are either too general with degraded classification ability or too specialized to generalize to unseen classes. In this paper, we propose an end-to-end network, termed as BGSNet, which equips and balances generalization and specialization abilities at the instance and dataset level. Specifically, BGSNet consists of two branches: the Generalization Network (GNet), which applies episodic meta-learning to learn generalized knowledge, and the Balanced Specialization Network (BSNet), which adopts multiple attentive extractors to extract discriminative features and achieve instance-level balance. A novel self-adjusted diversity loss is designed to optimize BSNet with redundancy reduced and diversity boosted. We further propose a differentiable dataset-level balance and update the weights in a linear annealing schedule to simulate network pruning and thus obtain the optimal structure for BSNet with dataset-level balance achieved. Experiments on four benchmark datasets demonstrate our model's effectiveness. Sufficient component ablations prove the necessity of integrating and balancing generalization and specialization abilities.
dc.language	en
dc.publisher	Institute of Electrical and Electronics Engineers (IEEE)
dc.relation.ispartof	IEEE Transactions on Multimedia
dc.relation.isbasedon	10.1109/TMM.2023.3236211
dc.rights	info:eu-repo/semantics/embargoedAccess
dc.subject	08 Information and Computing Sciences, 09 Engineering
dc.subject.classification	Artificial Intelligence & Image Processing
dc.title	Diversity-Boosted Generalization-Specialization Balancing for Zero-Shot Learning
dc.type	Journal Article
utslib.citation.volume	PP
utslib.for	08 Information and Computing Sciences
utslib.for	09 Engineering
pubs.organisational-group	/University of Technology Sydney
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology
pubs.organisational-group	/University of Technology Sydney/Strength - AAII - Australian Artificial Intelligence Institute
utslib.copyright.status	embargoed	*
utslib.copyright.embargo	2025-01-11T00:00:00+1000Z
dc.date.updated	2023-03-31T09:41:05Z
pubs.issue	99
pubs.publication-status	Published
pubs.volume	PP
utslib.citation.issue	99

Abstract:

Zero-Shot Learning (ZSL) aims to transfer classification capability from seen to unseen classes. Recent methods have proved that generalization and specialization are two essential abilities to achieve good performance in ZSL. However, focusing on only one of the abilities may result in models that are either too general with degraded classification ability or too specialized to generalize to unseen classes. In this paper, we propose an end-to-end network, termed as BGSNet, which equips and balances generalization and specialization abilities at the instance and dataset level. Specifically, BGSNet consists of two branches: the Generalization Network (GNet), which applies episodic meta-learning to learn generalized knowledge, and the Balanced Specialization Network (BSNet), which adopts multiple attentive extractors to extract discriminative features and achieve instance-level balance. A novel self-adjusted diversity loss is designed to optimize BSNet with redundancy reduced and diversity boosted. We further propose a differentiable dataset-level balance and update the weights in a linear annealing schedule to simulate network pruning and thus obtain the optimal structure for BSNet with dataset-level balance achieved. Experiments on four benchmark datasets demonstrate our model's effectiveness. Sufficient component ablations prove the necessity of integrating and balancing generalization and specialization abilities.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/168969