Comparison of Full-Reference Image Quality Models for Optimization of Image Processing Systems

The performance of objective image quality assessment (IQA) models has been evaluated primarily by comparing model predictions to human quality judgments. Perceptual datasets gathered for this purpose have provided useful benchmarks for improving IQA methods, but their heavy use creates a risk of ov...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:International journal of computer vision Ročník 129; číslo 4; s. 1258 - 1281
Hlavní autoři: Ding, Keyan, Ma, Kede, Wang, Shiqi, Simoncelli, Eero P.
Médium: Journal Article
Jazyk:angličtina
Vydáno: New York Springer US 01.04.2021
Springer
Springer Nature B.V
Témata:
Algorithms
Artificial Intelligence
Artificial neural networks
Compression tests
Computer Imaging
Computer Science
Equipment and supplies
Image compression
Image processing
Image Processing and Computer Vision
Image quality
Neural networks
Noise reduction
Optimization
Pattern Recognition
Pattern Recognition and Graphics
Quality assessment
Special Issue on Performance Evaluation in Computer Vision
Vision
Performance evaluation
Image quality assessment
Perceptual optimization
ISSN:0920-5691, 1573-1405
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:The performance of objective image quality assessment (IQA) models has been evaluated primarily by comparing model predictions to human quality judgments. Perceptual datasets gathered for this purpose have provided useful benchmarks for improving IQA methods, but their heavy use creates a risk of overfitting. Here, we perform a large-scale comparison of IQA models in terms of their use as objectives for the optimization of image processing algorithms. Specifically, we use eleven full-reference IQA models to train deep neural networks for four low-level vision tasks: denoising, deblurring, super-resolution, and compression. Subjective testing on the optimized images allows us to rank the competing models in terms of their perceptual performance, elucidate their relative advantages and disadvantages in these tasks, and propose a set of desirable properties for incorporation into future IQA models.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
Communicated by Daniel Scharstein.
ISSN:0920-5691
1573-1405
DOI:10.1007/s11263-020-01419-7