Scipediacontent: Scipediacontent moved page Draft Content 666839986 to Memon Korus 2018a

2021-02-01T20:06:39Z

Scipediacontent moved page Draft Content 666839986 to Memon Korus 2018a

Scipediacontent: Created page with " == Abstract == Forensic analysis of digital photo provenance relies on intrinsic traces left in the photograph at the time of its acquisition. Such analysis becomes unreliab..."

2021-02-01T20:06:35Z

Created page with " == Abstract == Forensic analysis of digital photo provenance relies on intrinsic traces left in the photograph at the time of its acquisition. Such analysis becomes unreliab..."

New page

== Abstract ==

Forensic analysis of digital photo provenance relies on intrinsic traces left in the photograph at the time of its acquisition. Such analysis becomes unreliable after heavy post-processing, such as down-sampling and re-compression applied upon distribution in the Web. This paper explores end-to-end optimization of the entire image acquisition and distribution workflow to facilitate reliable forensic analysis at the end of the distribution channel. We demonstrate that neural imaging pipelines can be trained to replace the internals of digital cameras, and jointly optimized for high-fidelity photo development and reliable provenance analysis. In our experiments, the proposed approach increased image manipulation detection accuracy from 45% to over 90%. The findings encourage further research towards building more reliable imaging pipelines with explicit provenance-guaranteeing properties.

Comment: Camera ready + supplement, CVPR'19

== Original document ==

The different versions of the original document can be found in:

* [http://arxiv.org/abs/1812.01516 http://arxiv.org/abs/1812.01516]

* [http://arxiv.org/pdf/1812.01516 http://arxiv.org/pdf/1812.01516]

* [http://xplorestaging.ieee.org/ielx7/8938205/8953184/08953353.pdf?arnumber=8953353 http://xplorestaging.ieee.org/ielx7/8938205/8953184/08953353.pdf?arnumber=8953353],
: [http://dx.doi.org/10.1109/cvpr.2019.00882 http://dx.doi.org/10.1109/cvpr.2019.00882]

* [https://dblp.uni-trier.de/db/conf/cvpr/cvpr2019.html#KorusM19 https://dblp.uni-trier.de/db/conf/cvpr/cvpr2019.html#KorusM19],
: [https://nyuscholars.nyu.edu/en/publications/content-authentication-for-neural-imaging-pipelines-end-to-end-op https://nyuscholars.nyu.edu/en/publications/content-authentication-for-neural-imaging-pipelines-end-to-end-op],
: [https://openaccess.thecvf.com/content_CVPR_2019/html/Korus_Content_Authentication_for_Neural_Imaging_Pipelines_End-To-End_Optimization_of_Photo_CVPR_2019_paper.html https://openaccess.thecvf.com/content_CVPR_2019/html/Korus_Content_Authentication_for_Neural_Imaging_Pipelines_End-To-End_Optimization_of_Photo_CVPR_2019_paper.html],
: [https://ui.adsabs.harvard.edu/abs/2018arXiv181201516K/abstract https://ui.adsabs.harvard.edu/abs/2018arXiv181201516K/abstract],
: [https://academic.microsoft.com/#/detail/2982544354 https://academic.microsoft.com/#/detail/2982544354]

← Older revision	Revision as of 20:06, 1 February 2021
(No difference)

Memon Korus 2018a - Revision history

Scipediacontent: Scipediacontent moved page Draft Content 666839986 to Memon Korus 2018a

Scipediacontent: Created page with " == Abstract == Forensic analysis of digital photo provenance relies on intrinsic traces left in the photograph at the time of its acquisition. Such analysis becomes unreliab..."