在現代社會之中，網路新聞、社群媒體以及各式的論壇已經成為我們生活之中不可或缺的一部分。也因此，如何辨別網路消息的真實性已經成為現代人相當重要的課題。過去，報章媒體講究有圖為證，圖片可以作為一篇報導的重要證明。但如今，隨著深度學習、生成對抗網路等技術的成熟，網路上也充斥著各式經過偽造的圖片。
生成對抗網路為一種非監督式學習的方法，透過讓兩個神經網路相互競爭對抗的方式來進行學習。生成對抗網路是由一個生成網路及一個辨別網路所組成，生成網路輸出的結果需要盡量接近真實樣本，而辨別網路則需要將生成網路所產生的虛假樣本從真實樣本中分辨出來。透過兩個網路不斷地相互競爭對抗、調整，從而產生人眼也無法區別、以假亂真的圖片。
本篇論文中，講述透過捲積神經網路、特徵萃取，以及透過遷移學習的方式來設計並訓練一個模型來對生成對抗網路所產生的偽造圖片進行辨識。相較於過去針對特定目標的訓練，透過遷移學習的方式可以大幅地降低訓練所需要的時間及運算成本，亦能獲得不錯的辨識度。

In this thesis, we study the detection of the authenticity of images generated by generative adversarial networks through cnn, feature extraction, and transfer learning. The problem of class imbalance is solved by data augmentation. Transfer learning is used to design and train a model that can be used to identify fake pictures. The time and computational cost required for training can be significantly reduced compared to traditional training methods, and good performance can be obtained.

[1] Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. Generative adversarial nets. In Advances in neural information processing systems, pages 2672–2680, 2014.
[2] Francesco Marra, Diego Gragnaniello, Davide Cozzolino, and Luisa Verdoliva. Detection of gan generated fake images over social networks. In 2018 IEEE Conference on Multimedia Information Processing and Retrieval (MIPR), pages 384–389. IEEE, 2018.
[3] Jun-Yan Zhu, Taesung Park, Phillip Isola, and Alexei A Efros. Unpaired image-to-image translation using cycle-consistent adversarial networks. In Proceedings of the IEEE international conference on computer vision, pages 2223–2232,2017.
[4] Lakshmanan Nataraj, Tajuddin Manhar Mohammed, BS Manjunath, Shivkumar Chandrasekaran, Arjuna Flenner, Jawadul H Bappy, and Amit K Roy-Chowdhury. Detecting gan generated fake images using cooccurrence matrices. Electronic Imaging, 2019(5):532–1, 2019.
[5] Scott McCloskey and Michael Albright. Detecting gan-generated imagery using color cues. arXiv preprint arXiv:1812.08247, 2018.
[6] Xu Zhang, Svebor Karaman, and Shih-Fu Chang. Detecting and simulating artifacts in gan fake images. In 2019 IEEE International Workshop on Information Forensics and Security (WIFS), pages 1–6. IEEE, 2019.
[7] Ian Goodfellow. Nips 2016 tutorial: Generative adversarial networks. arXiv preprint arXiv:1701.00160, 2016.
[8] Tim Salimans, Ian Goodfellow, Wojciech Zaremba, Vicki Cheung, Alec Rad ford, and Xi Chen. Improved techniques for training gans. In Advances in neural information processing systems, pages 2234–2242, 2016.
[9] Martin Heusel, Hubert Ramsauer, Thomas Unterthiner, Bernhard Nessler, and Sepp Hochreiter. Gans trained by a two time-scale update rule converge to a local nash equilibrium. In Advances in neural information processing systems, pages 6626–6637, 2017.
[10] Shane Barratt and Rishi Sharma. A note on the inception score. arXiv preprint arXiv:1801.01973, 2018.
[11] Christian Szegedy, Wei Liu, Yangqing Jia, Pierre Sermanet, Scott Reed, Dragomir Anguelov, Dumitru Erhan, Vincent Vanhoucke, and Andrew Rabinovich. Going deeper with convolutions. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 1–9, 2015.
[12] Fei-Fei Li. Imagenet. In http://image-net.org, 2009.
[13] Mehdi Mirza and Simon Osindero. Conditional generative adversarial nets. arXiv preprint arXiv:1411.1784, 2014.
[14] Yunjey Choi, Minje Choi, Munyoung Kim, Jung-Woo Ha, Sunghun Kim, and Jaegul Choo. Stargan: Unified generative adversarial networks for multi-domain image-to-image translation. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 8789–8797, 2018.
[15] Yann LeCun, Léon Bottou, Yoshua Bengio, and Patrick Haffner. Gradient based learning applied to document recognition. Proceedings of the IEEE, 86(11):2278–2324, 1998.
[16] Alex Krizhevsky, Ilya Sutskever, and Geoffrey E Hinton. Imagenet classification with deep convolutional neural networks. In Advances in neural information processing systems, pages 1097–1105, 2012.
[17] Alec Radford, Luke Metz, and Soumith Chintala. Unsupervised representation learning with deep convolutional generative adversarial networks. arXiv preprint arXiv:1511.06434, 2015.
[18] Xudong Mao, Qing Li, Haoran Xie, Raymond YK Lau, Zhen Wang, and Stephen Paul Smolley. Least squares generative adversarial networks. In Proceedings of the IEEE international conference on computer vision, pages 2794–2802, 2017.
[19] Martin Arjovsky, Soumith Chintala, and Léon Bottou. Wasserstein gan. arXiv preprint arXiv:1701.07875, 2017.
[20] Augustus Odena, Christopher Olah, and Jonathon Shlens. Conditional image synthesis with auxiliary classifier gans. In International conference on machine learning, pages 2642–2651, 2017.
[21] Jeff Donahue, Philipp Krähenbühl, and Trevor Darrell. Adversarial feature learning. arXiv preprint arXiv:1605.09782, 2016.
[22] François Chollet. Xception: Deep learning with depthwise separable convolutions. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 1251–1258, 2017.
[23] Gao Huang, Zhuang Liu, Laurens Van Der Maaten, and Kilian Q Weinberger. Densely connected convolutional networks. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 4700–4708, 2017.
[24] Sinno Jialin Pan and Qiang Yang. A survey on transfer learning. IEEE Transactions on knowledge and data engineering, 22(10):1345–1359, 2009.
[25] Run Wang, Felix Juefei-Xu, Lei Ma, Xiaofei Xie, Yihao Huang, Jian Wang, and Yang Liu. Fakespotter: A simple yet robust baseline for spotting ai-synthesized fake faces. In International Joint Conference on Artificial Intelligence (IJCAI), 2020.
[26] Brian Dolhansky, Joanna Bitton, Ben Pflaum, Jikuo Lu, Russ Howes, Menglin Wang, and Cristian Canton Ferrer. The deepfake detection challenge dataset. arXiv preprint arXiv:2006.07397, 2020.