Quality Assessment for Text-to-Image Generation: A Survey

Yu TIAN; Yue LIU; Shiqi WANG; Sam KWONG

doi:10.1109/MMUL.2025.3538862

Quality Assessment for Text-to-Image Generation: A Survey

Yu TIAN
, Yue LIU
, Shiqi WANG
, Sam KWONG

Lingnan University

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

6 Citations (Scopus)

Abstract

In recent years, there have been notable advancements in text-to-image generation facilitated by artificial intelligence (AI) technology. Text-to-image generation requires higher-level cognitive abilities, posing unique challenges for image quality assessment typically designed for professionally generated content and user-generated content. Existing works have extensively investigated quality assessment from subjective and objective perspectives, covering a range of evaluation dimensions such as text–image alignment, perception, esthetics, fairness, and toxicity. This article provides a comprehensive overview of recent advancements in image quality assessment for text-to-image generation. In particular, we review existing quality assessment studies from subjective and objective perspectives, highlighting representative datasets and objective metrics for assessing different aspects of AI-generated image quality. Additionally, we discuss the limitations of current research and propose future directions.

Original language	English
Pages (from-to)	44-52
Number of pages	9
Journal	IEEE Multimedia
Volume	32
Issue number	2
DOIs	https://doi.org/10.1109/MMUL.2025.3538862
Publication status	Published - 6 Feb 2025

Bibliographical note

Publisher Copyright:
© 1994-2012 IEEE.

Funding

This work was supported in part by the Science and Technology Innovation 2030 Key Project under Grant 2018AAA0101301, in part by the Hong Kong Innovation and Technology Commission (InnoHK Project CIM-DA), and in part by the Hong Kong General Research Fund under Grants 11209819, 11203820, 11200323, and 11203220. This article has supplementary down-loadable material available at https://doi.org/10.1109/MMUL.2025.3538862, provided by the authors.

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 9 Industry, Innovation, and Infrastructure

Access to Document

10.1109/MMUL.2025.3538862Licence: CC BY

https://ieeexplore.ieee.org/document/10877350/

Cite this

@article{7e31cc5bb35a43bd93ecfeeee88282d2,

title = "Quality Assessment for Text-to-Image Generation: A Survey",

abstract = "In recent years, there have been notable advancements in text-to-image generation facilitated by artificial intelligence (AI) technology. Text-to-image generation requires higher-level cognitive abilities, posing unique challenges for image quality assessment typically designed for professionally generated content and user-generated content. Existing works have extensively investigated quality assessment from subjective and objective perspectives, covering a range of evaluation dimensions such as text–image alignment, perception, esthetics, fairness, and toxicity. This article provides a comprehensive overview of recent advancements in image quality assessment for text-to-image generation. In particular, we review existing quality assessment studies from subjective and objective perspectives, highlighting representative datasets and objective metrics for assessing different aspects of AI-generated image quality. Additionally, we discuss the limitations of current research and propose future directions.",

author = "Yu TIAN and Yue LIU and Shiqi WANG and Sam KWONG",

note = "Publisher Copyright: {\textcopyright} 1994-2012 IEEE.",

year = "2025",

month = feb,

day = "6",

doi = "10.1109/MMUL.2025.3538862",

language = "English",

volume = "32",

pages = "44--52",

journal = "IEEE Multimedia",

issn = "1070-986X",

publisher = "IEEE Computer Society",

number = "2",

}

TY - JOUR

T1 - Quality Assessment for Text-to-Image Generation: A Survey

AU - TIAN, Yu

AU - LIU, Yue

AU - WANG, Shiqi

AU - KWONG, Sam

PY - 2025/2/6

Y1 - 2025/2/6

N2 - In recent years, there have been notable advancements in text-to-image generation facilitated by artificial intelligence (AI) technology. Text-to-image generation requires higher-level cognitive abilities, posing unique challenges for image quality assessment typically designed for professionally generated content and user-generated content. Existing works have extensively investigated quality assessment from subjective and objective perspectives, covering a range of evaluation dimensions such as text–image alignment, perception, esthetics, fairness, and toxicity. This article provides a comprehensive overview of recent advancements in image quality assessment for text-to-image generation. In particular, we review existing quality assessment studies from subjective and objective perspectives, highlighting representative datasets and objective metrics for assessing different aspects of AI-generated image quality. Additionally, we discuss the limitations of current research and propose future directions.

AB - In recent years, there have been notable advancements in text-to-image generation facilitated by artificial intelligence (AI) technology. Text-to-image generation requires higher-level cognitive abilities, posing unique challenges for image quality assessment typically designed for professionally generated content and user-generated content. Existing works have extensively investigated quality assessment from subjective and objective perspectives, covering a range of evaluation dimensions such as text–image alignment, perception, esthetics, fairness, and toxicity. This article provides a comprehensive overview of recent advancements in image quality assessment for text-to-image generation. In particular, we review existing quality assessment studies from subjective and objective perspectives, highlighting representative datasets and objective metrics for assessing different aspects of AI-generated image quality. Additionally, we discuss the limitations of current research and propose future directions.

UR - https://www.scopus.com/pages/publications/85217567058

U2 - 10.1109/MMUL.2025.3538862

DO - 10.1109/MMUL.2025.3538862

M3 - Journal Article (refereed)

SN - 1070-986X

VL - 32

SP - 44

EP - 52

JO - IEEE Multimedia

JF - IEEE Multimedia

IS - 2

ER -

Quality Assessment for Text-to-Image Generation: A Survey

Abstract

Bibliographical note

Funding

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this