AI Betrays AI
An Exploration of Applying Machine Learning in Facial Recognition
DOI:
https://doi.org/10.47611/jsrhs.v13i4.8058Keywords:
Machine Learning, Artificial Intelligence, Facial Recognition, Computer ScienceAbstract
The rise of AI deepfakes following the launch of ChatGPT and its AI counterpart DALL-E has sparked fear that the boundary between real and fake can no longer be identified. In this study, it was found that machine learning algorithms can be reliably used to distinguish between real and AI-generated images of human faces when provided with high resolution 300x300-pixel images with an accuracy score of 99.07%. This paper will cover the findings of this study by reviewing the main ideas behind machine learning, supervised learning, and neural networks and then examining the application of these techniques to a binary classification problem involving image classification.
Downloads
References or Bibliography
Mitchell, T. M. (1997). Machine Learning. McGraw-Hill Science/Engineering/Math.
Burkov, A. (2019). The Hundred-page Machine Learning Book. Andriy Burkov, 1.
Lu, Z., Huang, D., Bai, L., Qu, J., Wu, C., Liu, X., & Ouyang, W. (2023). Seeing is not always believing: Benchmarking Human and Model Perception of AI-Generated Images. Advances in Neural Information Processing Systems, 36. https://doi.org/10.48550/arXiv.2304.13023
Dettmers, T. (2022). “Deep Learning in a Nutshell: Core Concepts.” NVIDIA Technical Blog. https://developer.nvidia.com/blog/deep-learning-nutshell-core-concepts/
Boulahia, H. (2021). Fake-Vs-Real-Faces (Hard). https://www.kaggle.com/datasets/hamzaboulahia/hardfakevsrealfaces
Mistol, M. (2023). UTKFace (Aligned & Cropped). https://www.kaggle.com/datasets/moritzm00/utkface-cropped
TensorFlow Developers. (2024). TensorFlow (v2.17.0). Zenodo. https://doi.org/10.5281/zenodo.12726004
Keras: The high-level API for TensorFlow. TensorFlow. (n.d.) https://www.tensorflow.org/guide/keras
Mao, M. (2024). Machine Learning Model on Real vs. Deep-Fakes. https://www.kaggle.com/code/mikem27/machine-learning-model-on-real-vs-deep-fakes/notebook
Published
How to Cite
Issue
Section
Copyright (c) 2024 Mike Mao; Prof. Goldsztein, Ms. Gilberti
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
Copyright holder(s) granted JSR a perpetual, non-exclusive license to distriute & display this article.
