No-Reference Image Quality Assessment with Multi-Scale Orderless Pooling of Deep Features

No-Reference Image Quality Assessment with Multi-Scale Orderless Pooling of Deep Features

The goal of no-reference image quality assessment (NR-IQA) is to evaluate their perceptual quality of digital images without using the distortion-free, pristine counterparts. NR-IQA is an important part of multimedia signal processing since digital images can undergo a wide variety of distortions du...

Full description

Bibliographic Details
Main Author: Domonkos Varga
Format: Article
Language:English
Published: MDPI AG 2021-07-01
Series:Journal of Imaging
Subjects:
no-reference image quality assessment
deep learning
convolutional neural networks
Online Access:https://www.mdpi.com/2313-433X/7/7/112

Internet

https://www.mdpi.com/2313-433X/7/7/112

Similar Items