Hierarchical Degradation-aware Network for Full-Reference Image Quality Assessment

Xuting LAN, Fan JIA, Xu ZHUANG, Xuekai WEI*, Jun LUO, Mingliang ZHOU, Sam KWONG*

*Corresponding author for this work

Research output: Journal PublicationsJournal Article (refereed)peer-review

Abstract

Full-Reference Image Quality Assessment (FR-IQA) algorithms excel in evaluating perceptual distortions by comparing reference and distorted images. However, as the severity and quantity of distortions in datasets increase, existing FR-IQA methods struggle to capture complex nonlinear perceptual features. This limitation results in reduced adaptability and inaccurate assessments for images with more severe or multiple distortions. Recognizing the importance of understanding image degradation mechanisms, we propose a novel hierarchical degradation-aware network (HDaN) method. First, by exploring the degradation mechanisms from the reference image to the distorted image, our degradation network matches distortions that align more closely with the human visual system (HVS). Next, we design a convertor to project the matched features into multiple spaces, creating multidimensional feature representations that more comprehensively capture the complexity of image distortions rather than being confined to a single feature space. Then, we calculate a similarity matrix between the distorted and mapped features, selecting the most similar (top-k) features for merging. Finally, a regression network maps the merged features to quality scores, providing the final quality prediction. The experimental results demonstrate that our proposed HDaN method outperforms traditional deep learning-based FR-IQA methods. Specifically, the HDaN shows higher PLCC and SROCC metrics on benchmark datasets, significantly improving over existing methods. Moreover, the method exhibits better adaptability to images with varying degrees and types of distortions, thereby greatly enhancing the overall performance of IQA.
Original languageEnglish
Article number121557
JournalInformation Sciences
Volume690
Early online date17 Oct 2024
DOIs
Publication statusPublished - Feb 2025

Bibliographical note

Publisher Copyright:
© 2024 Elsevier Inc.

Funding

This work was supported in part by the National Natural Science Foundation of China under Grant 62176027, in part by the Key Projects of Basic Strengthening Plan under Grant 2022-JCJQ-ZD-018-11, in part by Chongqing Talent under Grant cstc2024ycjh-bgzxm0082, in part by the Joint Equipment Pre Research and Key Fund Project of the Ministry of Education under Grant 8091B012207, in part by the Human Resources and Social Security Bureau Project of Chongqing under Grant cx2020073, in part by Guangdong Oppo Mobile Telecommunications Corporation Ltd. under Grant H20240164, and in part by the Central University Operating Expenses under Grant 2024CDTGF-044.

Keywords

  • Deep learning
  • Degradation network
  • Feature extraction
  • Image quality assessment
  • Similarity

Fingerprint

Dive into the research topics of 'Hierarchical Degradation-aware Network for Full-Reference Image Quality Assessment'. Together they form a unique fingerprint.

Cite this