Show simple item record

dc.contributor.authorFan, Chunling
dc.contributor.authorLin, Hanhe
dc.contributor.authorHosu, Vlad
dc.contributor.authorZhang, Yun
dc.contributor.authorJiang, Qingshan
dc.contributor.authorHamzaoui, Raouf
dc.contributor.authorSaupe, Dietmar
dc.date.accessioned2019-03-26T15:29:05Z
dc.date.available2019-03-26T15:29:05Z
dc.date.issued2019-06
dc.identifier.citationC. Fan, H. Lin, V. Hosu, Y. Zhang, Q. Jiang, R. Hamzaoui, D. Saupe, (2019) SUR-Net: Predicting the satisfied user ratio curve for image compression with deep learning. In: Proc. 11th International Conference on Quality of Multimedia Experience (QoMEX), Berlin, June 2019.en
dc.identifier.urihttps://www.dora.dmu.ac.uk/handle/2086/17646
dc.descriptionThe file attached to this record is the author's final peer reviewed version. The Publisher's final version can be found by following the DOI link.en
dc.description.abstractThe Satisfied User Ratio (SUR) curve for a lossy image compression scheme, e.g., JPEG, characterizes the probability distribution of the Just Noticeable Difference (JND) level, the smallest distortion level that can be perceived by a subject. We propose the first deep learning approach to predict such SUR curves. Instead of the direct approach of regressing the SUR curve itself for a given reference image, our model is trained on pairs of images, original and compressed. Relying on a Siamese Convolutional Neural Network (CNN), feature pooling, a fully connected regression-head, and transfer learning, we achieved a good prediction performance. Experiments on the MCL-JCI dataset showed a mean Bhattacharyya distance between the predicted and the original JND distributions of only 0.072.en
dc.language.isoen_USen
dc.publisherIEEEen
dc.subjectSatisfied User Ratioen
dc.subjectJust Noticeable Differenceen
dc.subjectConvolutional Neural Networken
dc.subjectDeep Learningen
dc.titleSUR-Net: Predicting the Satisfied User Ratio Curve for Image Compression with Deep Learningen
dc.typeConferenceen
dc.identifier.doihttps://dx.doi.org/10.1109/QoMEX.2019.8743204
dc.peerreviewedYesen
dc.funderOther external funder (please detail below)en
dc.projectidNSFC Grant 61871372en
dc.projectidDFG project number 251654672 TRR 161en
dc.cclicenceCC-BY-NCen
dc.date.acceptance2019-02-26
dc.researchinstituteInstitute of Engineering Sciences (IES)en
dc.funder.otherNSFCen
dc.funder.otherDFGen


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record