English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

The mission of blind image quality evaluation is currently a challenging computer vision problem. Due to the shortage of reference images, it is hard for blind image quality evaluation methods to achieve the same performance as full reference image quality evaluation methods. In addition, current quality evaluation methods are difficult to effectively forecast the quality scores of synthesized distorted images as well as real distorted images. To address such issues, this study proposed a cyclic generative adversarial network composed of a quality perception network and a quality regression network on the ground of generative adversarial networks. For further enhancing the predictive performance of quality aware networks, this study proposed using attention blocks for adaptively fusing high-level semantic features and low-level semantic features. It extracted content and distortion information from images through an image quality evaluation method on the ground of content perception and distortion inference. And according to the different properties of the extracted features, adaptive fusion blocks were used for adaptively fusing content features and distortion features. Experiments showcased that the Spearman order correlation coefficient and Pearson linear correlation coefficient obtained by the proposed method on multiple datasets were higher than other similar methods. At the same time, the proposed method has achieved good prediction results on various types of distorted images, and has surpassed other methods. The prediction accuracy of the proposed method on five types of distortion was 0.971, 0.963, 0.984, 0.971, and 0.926, respectively. The proposed method achieved the highest predictive accuracy on all distortion types in the LIVE dataset, with predicted accuracy values of 0.973, 0.965, 0.984, 0.963, and 0.944, respectively. In summary, the proposed method not only achieved good prediction accuracy, but also had strong generalization performance in cross dataset testing. This provides a scientific and effective research direction for blind image quality evaluation.

Blind Image Quality Evaluation Method Based on Cyclic Generative Adversarial Network