A deep learning approach to image quality assessment
Deep convolutional neural networks (DCNNs) have an unchallengeable performance advantage over traditional machine learning in solving visual problems. However, DCNNs are vulnerable when the input signals are distorted or manipulated maliciously. We explore the computational modeling of image quality...
Saved in:
Main Author: | |
---|---|
Other Authors: | |
Format: | Theses and Dissertations |
Language: | English |
Published: |
2019
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/85493 http://hdl.handle.net/10220/50465 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
id |
sg-ntu-dr.10356-85493 |
---|---|
record_format |
dspace |
spelling |
sg-ntu-dr.10356-854932023-03-11T18:02:15Z A deep learning approach to image quality assessment Feng, Yeli Cai Yiyu School of Mechanical and Aerospace Engineering Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence Deep convolutional neural networks (DCNNs) have an unchallengeable performance advantage over traditional machine learning in solving visual problems. However, DCNNs are vulnerable when the input signals are distorted or manipulated maliciously. We explore the computational modeling of image quality assessment (IQA), investigate the vulnerability of DCNNs, and utilize IQA principle to mitigate it. Firstly, works that pushing the performance limit of IQA modeling is discussed. Using transfer learning, we re-purpose off-the-shelf visual features for quality prediction. The recurrent neural network is added to distill global features. Experiments show that the proposed IQA models outperform or perform on par with counterparts in the literature. Subsequently, a fit-for-task detection framework is introduced. Through exploiting the correlation of visual characteristics between maliciously manipulated images and conventional quality degradation, the detector effectively protect DCNNs from producing wrong results in scenarios of benign quality degradation and malicious attack. Likewise, radiographs of inadequate quality could lead to false diagnosis by image analysis systems. The image quality in medical radiology includes many more aspects beyond pixels. Lastly, we present a method that helps diagnosis AI to recognize view types of chest radiographs. Doctor of Philosophy 2019-11-25T06:47:27Z 2019-12-06T16:04:51Z 2019-11-25T06:47:27Z 2019-12-06T16:04:51Z 2019 Thesis Feng, Y. (2019). A deep learning approach to image quality assessment. Doctoral thesis, Nanyang Technological University, Singapore. https://hdl.handle.net/10356/85493 http://hdl.handle.net/10220/50465 10.32657/10356/85493 en 174 p. application/pdf |
institution |
Nanyang Technological University |
building |
NTU Library |
continent |
Asia |
country |
Singapore Singapore |
content_provider |
NTU Library |
collection |
DR-NTU |
language |
English |
topic |
Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence |
spellingShingle |
Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence Feng, Yeli A deep learning approach to image quality assessment |
description |
Deep convolutional neural networks (DCNNs) have an unchallengeable performance advantage over traditional machine learning in solving visual problems. However, DCNNs are vulnerable when the input signals are distorted or manipulated maliciously. We explore the computational modeling of image quality assessment (IQA), investigate the vulnerability of DCNNs, and utilize IQA principle to mitigate it. Firstly, works that pushing the performance limit of IQA modeling is discussed. Using transfer learning, we re-purpose off-the-shelf visual features for quality prediction. The recurrent neural network is added to distill global features. Experiments show that the proposed IQA models outperform or perform on par with counterparts in the literature. Subsequently, a fit-for-task detection framework is introduced. Through exploiting the correlation of visual characteristics between maliciously manipulated images and conventional quality degradation, the detector effectively protect DCNNs from producing wrong results in scenarios of benign quality degradation and malicious attack. Likewise, radiographs of inadequate quality could lead to false diagnosis by image analysis systems. The image quality in medical radiology includes many more aspects beyond pixels. Lastly, we present a method that helps diagnosis AI to recognize view types of chest radiographs. |
author2 |
Cai Yiyu |
author_facet |
Cai Yiyu Feng, Yeli |
format |
Theses and Dissertations |
author |
Feng, Yeli |
author_sort |
Feng, Yeli |
title |
A deep learning approach to image quality assessment |
title_short |
A deep learning approach to image quality assessment |
title_full |
A deep learning approach to image quality assessment |
title_fullStr |
A deep learning approach to image quality assessment |
title_full_unstemmed |
A deep learning approach to image quality assessment |
title_sort |
deep learning approach to image quality assessment |
publishDate |
2019 |
url |
https://hdl.handle.net/10356/85493 http://hdl.handle.net/10220/50465 |
_version_ |
1761782043782414336 |