Detecting fake facial images using deep machine learning
DOI:
https://doi.org/10.65204/djes.v3i2.664Keywords:
Image forensics, Deep fake, Metadata analysis, Generative Adversarial Network, Convolution Neural NetworksAbstract
One of the most popular phenomena is false photos that negatively affect our social lives, especially the sphere of politics and celeb. It is very easy to create fake images these days thanks to the pow-earful but easy to use apps available in a mobile device that navigate the social media network and because of the development of the Generative Adversarial Network (GAN), which may create pictures that are indistinguishable to the human eye. Which are fake pictures and fake videos simple to fake, hard to identify, simple to viralise. Consequently, image processing and artificial intelligence have a significant role to play in the resolution of such problems. Therefore, fake image detection is an urgent issue that should be managed and to avoid all these adverse outcomes. The proposed algorithm in this research was (Convolution Neural Network) that was the most popular algorithm used in deep learning to identify the fake images.
References
A picture’s worth, Digital Image Analysis and Forensics, N Krawetz - 2007 Ph D, Hacker Factor Solutions
http://imagej.net/Welcome ImageJ is an open source image processing program designed for scientific multidimensional images.J. Clerk Maxwell, A Treatise on Electricity and Magnetism, 3rd ed., vol. 2. Oxford: Clarendon, 1892, pp.68–73.
http://forensics.idealtest.org/ CASIA v2.0 CASIA V2.0 is with larger size and with more realistic and challenged fake images by using post-processing of tampered regions. It contains 7491 authentic and 5123 tampered color images.
http://neuroph.sourceforge.net/ Neuroph Framework Neuroph is lightweight Java neural network framework to develop common neural network architectures. It contains well designed, open source Java library with small number of basic classes which correspond to basic NN concepts.
https://github.com/drewnoakes/metadata-extractor Metadata-extractor is a straightforward Java library for reading metadata from image files.
G.Mohamed Sikandar, "100 Social Media Statistics You must know," [online] Available at:
https://blog.statusbrew.com/social-media statistics-2018-for-business/ [Accessed 02 Mar 2019].
Damian Radcliffe, Amanda Lam, "Social Media in the Middle East,"[online]Available:https://www.researchgate.net/publication/32318 5146_Social_Media_in_the_Middle_East_The_Story_of_2017 [Accessed 06 Feb 2019].
GMI_BLOGGER,"Saudi Arabia Social Media Statistics," GMI_ blogger. [online] Available at:https://www.globalmediainsight.com/ blog/saudi-arabia-social-media-statistics/ [Accessed 04 May 2019].
Kit Smith,"49 Incredible Instagram Statistics,". Brandwatch. [online] Available at: https://www.brandwatch.com/blog/instagram-stats/ [Accessed 10 May 2019].
Selling Stock. (2014). Selling Stock. [online] Available at: https://www. selling-stock.com/Article/18-billion-images-uploaded-to-the-web-every d [Accessed 12 Feb 2019].
Li, W., Prasad, S., Fowler, J. E., & Bruce, L. M. (2012). Locality preserving dimensionality reduction and classification for hyperspectral image analysis. IEEE Transactions on Geoscience and Remote Sensing, 50(4), 1185–1198.
A. Krizhevsky, I. Sutskever, & G. E. Hinton, (2012). Imagenet classification with deep convolutional neural networks. In Advances in Neural Information Processing Systems, 1097–1105.
K. Ravi, (2018). Detecting fake images with Machine Learning. Harkuch Journal [9] L. Zheng, Y. Yang, J. Zhang, Q. Cui, X. Zhang, Z. Li, et al. (2018). TI CNN: Convolutional Neural Networks for Fake News Detection. United
