Hybrid Deepfake Detection Utilizing MLP and LSTM

View on arXiv ← Back to list

Authors: Jacob Mallet, Natalie Krueger, Mounika Vanamala, Rushit Dave

Published: 2023-04-21 16:38:26+00:00

AI Summary

This research proposes a deepfake detection method using LSTM and MLP deep learning algorithms. The model was evaluated on the 140k Real and Fake Faces dataset, achieving an accuracy of up to 74.7% with the LSTM algorithm.

Abstract

The growing reliance of society on social media for authentic information has done nothing but increase over the past years. This has only raised the potential consequences of the spread of misinformation. One of the growing methods in popularity is to deceive users using a deepfake. A deepfake is an invention that has come with the latest technological advancements, which enables nefarious online users to replace their face with a computer generated, synthetic face of numerous powerful members of society. Deepfake images and videos now provide the means to mimic important political and cultural figures to spread massive amounts of false information. Models that can detect these deepfakes to prevent the spread of misinformation are now of tremendous necessity. In this paper, we propose a new deepfake detection schema utilizing two deep learning algorithms: long short term memory and multilayer perceptron. We evaluate our model using a publicly available dataset named 140k Real and Fake Faces to detect images altered by a deepfake with accuracies achieved as high as 74.7%

Key findings

The LSTM model outperformed the MLP model across all metrics, achieving a 74.7% accuracy. While the LSTM model showed higher recall (81.4%), the MLP model showed higher precision (69%). The findings highlight the potential of LSTM for deepfake image detection but also point to limitations in generalizability and real-world applicability.

Approach

The authors use a two-pronged approach, training separate models using LSTM and MLP architectures. Both models are trained on the same preprocessed dataset of real and fake images (from the 140k Real and Fake Faces dataset). Performance metrics including accuracy, precision, recall, F1 score, and AUC are used for evaluation.

Datasets

140k Real and Fake Faces dataset (combining Flickr-Faces-HQ and Deepfake Detection Challenge datasets)

Model(s)

Long Short-Term Memory (LSTM) and Multilayer Perceptron (MLP)

Author countries

USA

← Previous