A Review of Generalized Zero-Shot Learning Methods

Farhad POURPANAH; Moloud ABDAR; Yuxuan LUO; Xinlei ZHOU; Ran WANG; Chee Peng LIM; Xi-Zhao WANG; Q. M. Jonathan WU

doi:10.1109/TPAMI.2022.3191696

A Review of Generalized Zero-Shot Learning Methods

Farhad POURPANAH, Moloud ABDAR, Yuxuan LUO, Xinlei ZHOU, Ran WANG^*, Chee Peng LIM, Xi-Zhao WANG, Q. M. Jonathan WU

^*Corresponding author for this work

Research output: Journal Publications › Journal Article (refereed) › peer-review

112 Citations (Scopus)

Abstract

Generalized zero-shot learning (GZSL) aims to train a model for classifying data samples under the condition that some output classes are unknown during supervised learning. To address this challenging task, GZSL leverages semantic information of the seen (source) and unseen (target) classes to bridge the gap between both seen and unseen classes. Since its introduction, many GZSL models have been formulated. In this review paper, we present a comprehensive review on GZSL. First, we provide an overview of GZSL including the problems and challenges. Then, we introduce a hierarchical categorization for the GZSL methods and discuss the representative methods in each category. In addition, we discuss the available benchmark data sets and applications of GZSL, along with a discussion on the research gaps and directions for future investigations.

Original language	English
Pages (from-to)	4051-4070
Number of pages	20
Journal	IEEE Transactions on Pattern Analysis and Machine Intelligence
Volume	45
Issue number	4
Early online date	18 Jul 2022
DOIs	https://doi.org/10.1109/TPAMI.2022.3191696
Publication status	Published - 1 Apr 2023
Externally published	Yes

Bibliographical note

This work was supported in part by the National Natural Science Foundation of China under Grant 62176160, 61732011 and 61976141, in part by Guangdong Basic and Applied Basic Research Foundation under Grant 2022A1515010791, in part by the Natural Science Foundation of Shenzhen University Stability Support Program under Grant no. 20200804193857002, and in part by Interdisciplinary Innovation Team of SZU.

Keywords

deep learning
Generalized zero shot learning
generative adversarial networks
semantic embedding
variational auto-encoders

Access to Document

10.1109/TPAMI.2022.3191696Licence: CC BY-NC-ND

Cite this

@article{54d6d8603b364ef6af9c2c9f88030d0b,

title = "A Review of Generalized Zero-Shot Learning Methods",

abstract = "Generalized zero-shot learning (GZSL) aims to train a model for classifying data samples under the condition that some output classes are unknown during supervised learning. To address this challenging task, GZSL leverages semantic information of the seen (source) and unseen (target) classes to bridge the gap between both seen and unseen classes. Since its introduction, many GZSL models have been formulated. In this review paper, we present a comprehensive review on GZSL. First, we provide an overview of GZSL including the problems and challenges. Then, we introduce a hierarchical categorization for the GZSL methods and discuss the representative methods in each category. In addition, we discuss the available benchmark data sets and applications of GZSL, along with a discussion on the research gaps and directions for future investigations.",

keywords = "deep learning, Generalized zero shot learning, generative adversarial networks, semantic embedding, variational auto-encoders",

author = "Farhad POURPANAH and Moloud ABDAR and Yuxuan LUO and Xinlei ZHOU and Ran WANG and LIM, {Chee Peng} and Xi-Zhao WANG and WU, {Q. M. Jonathan}",

note = "This work was supported in part by the National Natural Science Foundation of China under Grant 62176160, 61732011 and 61976141, in part by Guangdong Basic and Applied Basic Research Foundation under Grant 2022A1515010791, in part by the Natural Science Foundation of Shenzhen University Stability Support Program under Grant no. 20200804193857002, and in part by Interdisciplinary Innovation Team of SZU.",

year = "2023",

month = apr,

day = "1",

doi = "10.1109/TPAMI.2022.3191696",

language = "English",

volume = "45",

pages = "4051--4070",

journal = "IEEE Transactions on Pattern Analysis and Machine Intelligence",

issn = "0162-8828",

publisher = "IEEE Computer Society",

number = "4",

}

TY - JOUR

T1 - A Review of Generalized Zero-Shot Learning Methods

AU - POURPANAH, Farhad

AU - ABDAR, Moloud

AU - LUO, Yuxuan

AU - ZHOU, Xinlei

AU - WANG, Ran

AU - LIM, Chee Peng

AU - WANG, Xi-Zhao

AU - WU, Q. M. Jonathan

N1 - This work was supported in part by the National Natural Science Foundation of China under Grant 62176160, 61732011 and 61976141, in part by Guangdong Basic and Applied Basic Research Foundation under Grant 2022A1515010791, in part by the Natural Science Foundation of Shenzhen University Stability Support Program under Grant no. 20200804193857002, and in part by Interdisciplinary Innovation Team of SZU.

PY - 2023/4/1

Y1 - 2023/4/1

N2 - Generalized zero-shot learning (GZSL) aims to train a model for classifying data samples under the condition that some output classes are unknown during supervised learning. To address this challenging task, GZSL leverages semantic information of the seen (source) and unseen (target) classes to bridge the gap between both seen and unseen classes. Since its introduction, many GZSL models have been formulated. In this review paper, we present a comprehensive review on GZSL. First, we provide an overview of GZSL including the problems and challenges. Then, we introduce a hierarchical categorization for the GZSL methods and discuss the representative methods in each category. In addition, we discuss the available benchmark data sets and applications of GZSL, along with a discussion on the research gaps and directions for future investigations.

AB - Generalized zero-shot learning (GZSL) aims to train a model for classifying data samples under the condition that some output classes are unknown during supervised learning. To address this challenging task, GZSL leverages semantic information of the seen (source) and unseen (target) classes to bridge the gap between both seen and unseen classes. Since its introduction, many GZSL models have been formulated. In this review paper, we present a comprehensive review on GZSL. First, we provide an overview of GZSL including the problems and challenges. Then, we introduce a hierarchical categorization for the GZSL methods and discuss the representative methods in each category. In addition, we discuss the available benchmark data sets and applications of GZSL, along with a discussion on the research gaps and directions for future investigations.

KW - deep learning

KW - Generalized zero shot learning

KW - generative adversarial networks

KW - semantic embedding

KW - variational auto-encoders

UR - http://www.scopus.com/inward/record.url?scp=85135235120&partnerID=8YFLogxK

U2 - 10.1109/TPAMI.2022.3191696

DO - 10.1109/TPAMI.2022.3191696

M3 - Journal Article (refereed)

C2 - 35849673

AN - SCOPUS:85135235120

SN - 0162-8828

VL - 45

SP - 4051

EP - 4070

JO - IEEE Transactions on Pattern Analysis and Machine Intelligence

JF - IEEE Transactions on Pattern Analysis and Machine Intelligence

IS - 4

ER -

A Review of Generalized Zero-Shot Learning Methods

Abstract

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this