Home > Research > Publications & Outputs > Detecting Post Editing of Multimedia Images usi...

Links

Text available via DOI:

View graph of relations

Detecting Post Editing of Multimedia Images using Transfer Learning and Fine Tuning

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Published

Standard

Detecting Post Editing of Multimedia Images using Transfer Learning and Fine Tuning. / Jonker, Simon; Jelstrup, Malthe; Meng, Weizhi et al.
In: ACM Transactions on Multimedia Computing, Communications, and Applications, Vol. 20, No. 6, 154, 30.06.2024.

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Harvard

Jonker, S, Jelstrup, M, Meng, W & Lampe, B 2024, 'Detecting Post Editing of Multimedia Images using Transfer Learning and Fine Tuning', ACM Transactions on Multimedia Computing, Communications, and Applications, vol. 20, no. 6, 154. https://doi.org/10.1145/3633284

APA

Jonker, S., Jelstrup, M., Meng, W., & Lampe, B. (2024). Detecting Post Editing of Multimedia Images using Transfer Learning and Fine Tuning. ACM Transactions on Multimedia Computing, Communications, and Applications, 20(6), Article 154. https://doi.org/10.1145/3633284

Vancouver

Jonker S, Jelstrup M, Meng W, Lampe B. Detecting Post Editing of Multimedia Images using Transfer Learning and Fine Tuning. ACM Transactions on Multimedia Computing, Communications, and Applications. 2024 Jun 30;20(6):154. Epub 2024 Mar 8. doi: 10.1145/3633284

Author

Jonker, Simon ; Jelstrup, Malthe ; Meng, Weizhi et al. / Detecting Post Editing of Multimedia Images using Transfer Learning and Fine Tuning. In: ACM Transactions on Multimedia Computing, Communications, and Applications. 2024 ; Vol. 20, No. 6.

Bibtex

@article{5e3e82d2e02b41f3aa21b016b2723464,
title = "Detecting Post Editing of Multimedia Images using Transfer Learning and Fine Tuning",
abstract = "In the domain of general image forgery detection, a myriad of different classification solutions have been developed to distinguish a “tampered” image from a “pristine” image. In this work, we aim to develop a new method to tackle the problem of binary image forgery detection. Our approach builds upon the extensive training that state-of-the-art image classification models have undergone on regular images from the ImageNet dataset, and transfers that knowledge to the image forgery detection space. By leveraging transfer learning and fine tuning, we can fit state-of-the-art image classification models to the forgery detection task. We train the models on a diverse and evenly distributed image forgery dataset. With five models—EfficientNetB0, VGG16, Xception, ResNet50V2, and NASNet-Large—we transferred and adapted pre-trained knowledge from ImageNet to the forgery detection task. Each model was fitted, fine-tuned, and evaluated according to a set of performance metrics. Our evaluation demonstrated the efficacy of large-scale image classification models—paired with transfer learning and fine tuning—at detecting image forgeries. When pitted against a previously unseen dataset, the best-performing model of EfficientNetB0 could achieve an accuracy rate of nearly 89.7%.",
author = "Simon Jonker and Malthe Jelstrup and Weizhi Meng and Brooke Lampe",
year = "2024",
month = jun,
day = "30",
doi = "10.1145/3633284",
language = "English",
volume = "20",
journal = "ACM Transactions on Multimedia Computing, Communications, and Applications",
issn = "1551-6857",
publisher = "Association for Computing Machinery (ACM)",
number = "6",

}

RIS

TY - JOUR

T1 - Detecting Post Editing of Multimedia Images using Transfer Learning and Fine Tuning

AU - Jonker, Simon

AU - Jelstrup, Malthe

AU - Meng, Weizhi

AU - Lampe, Brooke

PY - 2024/6/30

Y1 - 2024/6/30

N2 - In the domain of general image forgery detection, a myriad of different classification solutions have been developed to distinguish a “tampered” image from a “pristine” image. In this work, we aim to develop a new method to tackle the problem of binary image forgery detection. Our approach builds upon the extensive training that state-of-the-art image classification models have undergone on regular images from the ImageNet dataset, and transfers that knowledge to the image forgery detection space. By leveraging transfer learning and fine tuning, we can fit state-of-the-art image classification models to the forgery detection task. We train the models on a diverse and evenly distributed image forgery dataset. With five models—EfficientNetB0, VGG16, Xception, ResNet50V2, and NASNet-Large—we transferred and adapted pre-trained knowledge from ImageNet to the forgery detection task. Each model was fitted, fine-tuned, and evaluated according to a set of performance metrics. Our evaluation demonstrated the efficacy of large-scale image classification models—paired with transfer learning and fine tuning—at detecting image forgeries. When pitted against a previously unseen dataset, the best-performing model of EfficientNetB0 could achieve an accuracy rate of nearly 89.7%.

AB - In the domain of general image forgery detection, a myriad of different classification solutions have been developed to distinguish a “tampered” image from a “pristine” image. In this work, we aim to develop a new method to tackle the problem of binary image forgery detection. Our approach builds upon the extensive training that state-of-the-art image classification models have undergone on regular images from the ImageNet dataset, and transfers that knowledge to the image forgery detection space. By leveraging transfer learning and fine tuning, we can fit state-of-the-art image classification models to the forgery detection task. We train the models on a diverse and evenly distributed image forgery dataset. With five models—EfficientNetB0, VGG16, Xception, ResNet50V2, and NASNet-Large—we transferred and adapted pre-trained knowledge from ImageNet to the forgery detection task. Each model was fitted, fine-tuned, and evaluated according to a set of performance metrics. Our evaluation demonstrated the efficacy of large-scale image classification models—paired with transfer learning and fine tuning—at detecting image forgeries. When pitted against a previously unseen dataset, the best-performing model of EfficientNetB0 could achieve an accuracy rate of nearly 89.7%.

U2 - 10.1145/3633284

DO - 10.1145/3633284

M3 - Journal article

VL - 20

JO - ACM Transactions on Multimedia Computing, Communications, and Applications

JF - ACM Transactions on Multimedia Computing, Communications, and Applications

SN - 1551-6857

IS - 6

M1 - 154

ER -