KonIQ-10k: An Ecologically Valid Database for Deep Learning of Blind Image Quality Assessment

Deep learning methods for image quality assessment (IQA) are limited due to the small size of existing datasets. Extensive datasets require substantial resources both for generating publishable content and annotating it accurately. We present a systematic and scalable approach to creating KonIQ-10k,...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on image processing 2020-01, Vol.29, p.4041-4056
Main Authors: Hosu, Vlad, Lin, Hanhe, Sziranyi, Tamas, Saupe, Dietmar
Format: Article
Language:English
Subjects:
Biological system modeling
blind image quality assessment
convolutional neural networks
Correlation analysis
Crowdsourcing
Datasets
Deep learning
Distortion
diversity sampling
Ecological monitoring
Feature extraction
Image database
Image quality
Machine learning
Quality
Quality assessment
subjective image quality assessment
Training
User generated content
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Deep learning methods for image quality assessment (IQA) are limited due to the small size of existing datasets. Extensive datasets require substantial resources both for generating publishable content and annotating it accurately. We present a systematic and scalable approach to creating KonIQ-10k, the largest IQA dataset to date, consisting of 10,073 quality scored images. It is the first in-the-wild database aiming for ecological validity, concerning the authenticity of distortions, the diversity of content, and quality-related indicators. Through the use of crowdsourcing, we obtained 1.2 million reliable quality ratings from 1,459 crowd workers, paving the way for more general IQA models. We propose a novel, deep learning model (KonCept512), to show an excellent generalization beyond the test set (0.921 SROCC), to the current state-of-the-art database LIVE-in-the-Wild (0.825 SROCC). The model derives its core performance from the InceptionResNet architecture, being trained at a higher resolution than previous models (512 × 384 ). Correlation analysis shows that KonCept512 performs similar to having 9 subjective scores for each test image.
ISSN:1057-7149
1941-0042
DOI:10.1109/TIP.2020.2967829