A quality assessment algorithm for no-reference images based on transfer learning
Image quality assessment (IQA) plays a critical role in automatically detecting and correcting defects in images, thereby enhancing the overall performance of image processing and transmission systems. While research on reference-based IQA is well-established, studies on no-reference image IQA remai...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
PeerJ Inc.
2025-01-01
|
| Series: | PeerJ Computer Science |
| Subjects: | |
| Online Access: | https://peerj.com/articles/cs-2654.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832570596984094720 |
|---|---|
| author | Yang Yang Chang Liu Hui Wu Dingguo Yu |
| author_facet | Yang Yang Chang Liu Hui Wu Dingguo Yu |
| author_sort | Yang Yang |
| collection | DOAJ |
| description | Image quality assessment (IQA) plays a critical role in automatically detecting and correcting defects in images, thereby enhancing the overall performance of image processing and transmission systems. While research on reference-based IQA is well-established, studies on no-reference image IQA remain underdeveloped. In this article, we propose a novel no-reference IQA algorithm based on transfer learning (IQA-NRTL). This algorithm leverages a deep convolutional neural network (CNN) due to its ability to effectively capture multi-scale semantic information features, which are essential for representing the complex visual perception in images. These features are extracted through a visual perception module. Subsequently, an adaptive fusion network integrates these features, and a fully connected regression network correlates the fused semantic information with global semantic information to perform the final quality assessment. Experimental results on authentically distorted datasets (KonIQ-10k, BIQ2021), synthetically distorted datasets (LIVE, TID2013), and an artificial intelligence (AI)-generated content dataset (AGIQA-1K) show that the proposed IQA-NRTL algorithm significantly improves performance compared to mainstream no-reference IQA algorithms, depending on variations in image content and complexity. |
| format | Article |
| id | doaj-art-de6a7aaef2d64ba09ee83bb69e6a5e00 |
| institution | Kabale University |
| issn | 2376-5992 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | PeerJ Inc. |
| record_format | Article |
| series | PeerJ Computer Science |
| spelling | doaj-art-de6a7aaef2d64ba09ee83bb69e6a5e002025-02-02T15:05:20ZengPeerJ Inc.PeerJ Computer Science2376-59922025-01-0111e265410.7717/peerj-cs.2654A quality assessment algorithm for no-reference images based on transfer learningYang YangChang LiuHui WuDingguo YuImage quality assessment (IQA) plays a critical role in automatically detecting and correcting defects in images, thereby enhancing the overall performance of image processing and transmission systems. While research on reference-based IQA is well-established, studies on no-reference image IQA remain underdeveloped. In this article, we propose a novel no-reference IQA algorithm based on transfer learning (IQA-NRTL). This algorithm leverages a deep convolutional neural network (CNN) due to its ability to effectively capture multi-scale semantic information features, which are essential for representing the complex visual perception in images. These features are extracted through a visual perception module. Subsequently, an adaptive fusion network integrates these features, and a fully connected regression network correlates the fused semantic information with global semantic information to perform the final quality assessment. Experimental results on authentically distorted datasets (KonIQ-10k, BIQ2021), synthetically distorted datasets (LIVE, TID2013), and an artificial intelligence (AI)-generated content dataset (AGIQA-1K) show that the proposed IQA-NRTL algorithm significantly improves performance compared to mainstream no-reference IQA algorithms, depending on variations in image content and complexity.https://peerj.com/articles/cs-2654.pdfImage quality assessment (IQA)Non-reference image quality assessment (IQA-NRTL)Transfer learningDeep convolutional neural networkAdaptive fusion network |
| spellingShingle | Yang Yang Chang Liu Hui Wu Dingguo Yu A quality assessment algorithm for no-reference images based on transfer learning PeerJ Computer Science Image quality assessment (IQA) Non-reference image quality assessment (IQA-NRTL) Transfer learning Deep convolutional neural network Adaptive fusion network |
| title | A quality assessment algorithm for no-reference images based on transfer learning |
| title_full | A quality assessment algorithm for no-reference images based on transfer learning |
| title_fullStr | A quality assessment algorithm for no-reference images based on transfer learning |
| title_full_unstemmed | A quality assessment algorithm for no-reference images based on transfer learning |
| title_short | A quality assessment algorithm for no-reference images based on transfer learning |
| title_sort | quality assessment algorithm for no reference images based on transfer learning |
| topic | Image quality assessment (IQA) Non-reference image quality assessment (IQA-NRTL) Transfer learning Deep convolutional neural network Adaptive fusion network |
| url | https://peerj.com/articles/cs-2654.pdf |
| work_keys_str_mv | AT yangyang aqualityassessmentalgorithmfornoreferenceimagesbasedontransferlearning AT changliu aqualityassessmentalgorithmfornoreferenceimagesbasedontransferlearning AT huiwu aqualityassessmentalgorithmfornoreferenceimagesbasedontransferlearning AT dingguoyu aqualityassessmentalgorithmfornoreferenceimagesbasedontransferlearning AT yangyang qualityassessmentalgorithmfornoreferenceimagesbasedontransferlearning AT changliu qualityassessmentalgorithmfornoreferenceimagesbasedontransferlearning AT huiwu qualityassessmentalgorithmfornoreferenceimagesbasedontransferlearning AT dingguoyu qualityassessmentalgorithmfornoreferenceimagesbasedontransferlearning |