Image Quality Assessment: Measuring Perceptual Degradation via Distribution Measures in Deep Feature Spaces

Xingran LIAO, Xuekai WEI, Mingliang ZHOU*, Zhengguo LI, Sam KWONG*

*Corresponding author for this work

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


This study aims to develop advanced and training-free full-reference image quality assessment (FR-IQA) models based on deep neural networks. Specifically, we investigate measures that allow us to perceptually compare deep network features and reveal their underlying factors. We find that distribution measures enjoy advanced perceptual awareness and test the Wasserstein distance (WSD), Jensen-Shannon divergence (JSD), and symmetric Kullback-Leibler divergence (SKLD) measures when comparing deep features acquired from various pretrained deep networks, including the Visual Geometry Group (VGG) network, SqueezeNet, MobileNet, and EfficientNet. The proposed FR-IQA models exhibit superior alignment with subjective human evaluations across diverse image quality assessment (IQA) datasets without training, demonstrating the advanced perceptual relevance of distribution measures when comparing deep network features. Additionally, we explore the applicability of deep distribution measures in image super-resolution enhancement tasks, highlighting their potential for guiding perceptual enhancements. The code is available on website. (

Original languageEnglish
Pages (from-to)4044-4059
Number of pages16
JournalIEEE Transactions on Image Processing
Early online date28 Jun 2024
Publication statusPublished - 2024

Bibliographical note

Publisher Copyright:


  • Image quality assessment
  • deep neural network image representations
  • distribution measures
  • perceptual degradation
  • perceptual optimization
  • training-free


Dive into the research topics of 'Image Quality Assessment: Measuring Perceptual Degradation via Distribution Measures in Deep Feature Spaces'. Together they form a unique fingerprint.

Cite this