eduzhai > Applied Sciences > Engineering >

No-Reference Image Quality Assessment via Feature Fusion and Multi-Task Learning

  • Save

... pages left unread,continue reading

Document pages: 10 pages

Abstract: Blind or no-reference image quality assessment (NR-IQA) is a fundamental,unsolved, and yet challenging problem due to the unavailability of a referenceimage. It is vital to the streaming and social media industries that impactbillions of viewers daily. Although previous NR-IQA methods leveraged differentfeature extraction approaches, the performance bottleneck still exists. In thispaper, we propose a simple and yet effective general-purpose no-reference (NR)image quality assessment (IQA) framework based on multi-task learning. Ourmodel employs distortion types as well as subjective human scores to predictimage quality. We propose a feature fusion method to utilize distortioninformation to improve the quality score estimation task. In our experiments,we demonstrate that by utilizing multi-task learning and our proposed featurefusion method, our model yields better performance for the NR-IQA task. Todemonstrate the effectiveness of our approach, we test our approach on sevenstandard datasets and show that we achieve state-of-the-art results on variousdatasets.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×