Evaluation of Pre-Trained CNN Models for Geographic Fake Image Detection

View on arXiv ← Back to list

Authors: Sid Ahmed Fezza, Mohammed Yasser Ouis, Bachir Kaddar, Wassim Hamidouche, Abdenour Hadid

Published: 2022-10-01 20:37:24+00:00

AI Summary

This research paper benchmarks four pre-trained Convolutional Neural Network (CNN) architectures for detecting fake satellite images. The authors evaluate their performance and robustness against image distortions, establishing new baselines for this under-explored area of deepfake detection.

Abstract

Thanks to the remarkable advances in generative adversarial networks (GANs), it is becoming increasingly easy to generate/manipulate images. The existing works have mainly focused on deepfake in face images and videos. However, we are currently witnessing the emergence of fake satellite images, which can be misleading or even threatening to national security. Consequently, there is an urgent need to develop detection methods capable of distinguishing between real and fake satellite images. To advance the field, in this paper, we explore the suitability of several convolutional neural network (CNN) architectures for fake satellite image detection. Specifically, we benchmark four CNN models by conducting extensive experiments to evaluate their performance and robustness against various image distortions. This work allows the establishment of new baselines and may be useful for the development of CNN-based methods for fake satellite image detection.

Key findings

ResNet-50 achieved the best performance with an F1 score of 0.990. CNN-based methods significantly outperformed methods using handcrafted features. Image distortions, especially Gaussian noise, negatively impacted detection performance.

Approach

The paper uses transfer learning. Four pre-trained CNN models (VGG16, ResNet50, InceptionV3, and Xception) are fine-tuned on a dataset of real and fake satellite images. The last fully connected layer is removed and replaced with three new fully connected layers for classification.

Datasets

UW fake satellite image dataset

Model(s)

VGG16, ResNet50, InceptionV3, Xception

Author countries

Algeria, France, UAE

← Previous