Image Forgery Detection and Localization Using a Fully Convolutional Network

David Fischinger; David Schreiber; Martin Boyer

Image Forgery Detection and Localization Using a Fully Convolutional Network

David Fischinger (Speaker)
, David Schreiber
, Martin Boyer

Research output: Chapter in Book or Conference Proceedings › Conference Proceedings with Oral Presentation › peer-review

Abstract

To fight the growing problem of fake news – and specifically image manipulation – we propose a simple, yet efficient neural network architecture for detecting and localizing various image forgeries on a pixel-level. Robust features for forgery detection and localization were learned and the trained model performs well, even on heavily downscaled images, but without the excessive processing time of
competitive approaches based on image decomposition and merging of the fragmental results. We provide detailed explanations regarding the creation of our training dataset
comprising 1.9 million images. Finally, we compare the proposed solution against several state-of-the-art methods on four public benchmark datasets in order to demonstrate
its superior performance.

Original language	English
Title of host publication	Proceedings of the OAGM Workshop 2022
Subtitle of host publication	Digitalization for Smart Farming and Forestry
Editors	Hermann Bürstmayr, Andreas Gronauer, Andreas Holzinger, Peter M. Roth, Karl Stampfer
Pages	19-25
Number of pages	7
Publication status	Published - 2023
Event	OAGM Workshop 2022 - Tulln, Austria Duration: 7 Nov 2022 → 8 Nov 2022

Workshop

Workshop	OAGM Workshop 2022
Country/Territory	Austria
Period	7/11/22 → 8/11/22

Research Field

Former Research Field - Surveillance and Protection

Access to Document

https://workshops.aapr.at/Documents/proceedings.pdf

Cite this

@inproceedings{0c9a0472aa694d6ea908a07a50139765,

title = "Image Forgery Detection and Localization Using a Fully Convolutional Network",

abstract = "To fight the growing problem of fake news – and specifically image manipulation – we propose a simple, yet efficient neural network architecture for detecting and localizing various image forgeries on a pixel-level. Robust features for forgery detection and localization were learned and the trained model performs well, even on heavily downscaled images, but without the excessive processing time of competitive approaches based on image decomposition and merging of the fragmental results. We provide detailed explanations regarding the creation of our training dataset comprising 1.9 million images. Finally, we compare the proposed solution against several state-of-the-art methods on four public benchmark datasets in order to demonstrate its superior performance.",

author = "David Fischinger and David Schreiber and Martin Boyer",

year = "2023",

language = "English",

pages = "19--25",

editor = "B{\"u}rstmayr, \{Hermann \} and Andreas Gronauer and Holzinger, \{Andreas \} and Roth, \{Peter M.\} and Stampfer, \{Karl \}",

booktitle = "Proceedings of the OAGM Workshop 2022",

note = "OAGM Workshop 2022 ; Conference date: 07-11-2022 Through 08-11-2022",

}

Fischinger, D, Schreiber, D & Boyer, M 2023, Image Forgery Detection and Localization Using a Fully Convolutional Network. in H Bürstmayr, A Gronauer, A Holzinger, PM Roth & K Stampfer (eds), Proceedings of the OAGM Workshop 2022: Digitalization for Smart Farming and Forestry. pp. 19-25, OAGM Workshop 2022, Austria, 7/11/22. <https://workshops.aapr.at/Documents/proceedings.pdf>

Image Forgery Detection and Localization Using a Fully Convolutional Network. / Fischinger, David (Speaker); Schreiber, David; Boyer, Martin.
Proceedings of the OAGM Workshop 2022: Digitalization for Smart Farming and Forestry. ed. / Hermann Bürstmayr; Andreas Gronauer; Andreas Holzinger; Peter M. Roth; Karl Stampfer. 2023. p. 19-25.

Research output: Chapter in Book or Conference Proceedings › Conference Proceedings with Oral Presentation › peer-review

TY - GEN

T1 - Image Forgery Detection and Localization Using a Fully Convolutional Network

AU - Schreiber, David

AU - Boyer, Martin

A2 - Fischinger, David

A2 - Bürstmayr, Hermann

A2 - Gronauer, Andreas

A2 - Holzinger, Andreas

A2 - Roth, Peter M.

A2 - Stampfer, Karl

PY - 2023

Y1 - 2023

N2 - To fight the growing problem of fake news – and specifically image manipulation – we propose a simple, yet efficient neural network architecture for detecting and localizing various image forgeries on a pixel-level. Robust features for forgery detection and localization were learned and the trained model performs well, even on heavily downscaled images, but without the excessive processing time of competitive approaches based on image decomposition and merging of the fragmental results. We provide detailed explanations regarding the creation of our training dataset comprising 1.9 million images. Finally, we compare the proposed solution against several state-of-the-art methods on four public benchmark datasets in order to demonstrate its superior performance.

AB - To fight the growing problem of fake news – and specifically image manipulation – we propose a simple, yet efficient neural network architecture for detecting and localizing various image forgeries on a pixel-level. Robust features for forgery detection and localization were learned and the trained model performs well, even on heavily downscaled images, but without the excessive processing time of competitive approaches based on image decomposition and merging of the fragmental results. We provide detailed explanations regarding the creation of our training dataset comprising 1.9 million images. Finally, we compare the proposed solution against several state-of-the-art methods on four public benchmark datasets in order to demonstrate its superior performance.

M3 - Conference Proceedings with Oral Presentation

SP - 19

EP - 25

BT - Proceedings of the OAGM Workshop 2022

T2 - OAGM Workshop 2022

Y2 - 7 November 2022 through 8 November 2022

ER -

Image Forgery Detection and Localization Using a Fully Convolutional Network

Abstract

Workshop

Research Field

Access to Document

Fingerprint

Cite this