Verify It Yourself: A Note on Activation Functions’ Influence on Fast DeepFake Detection

Piotr Kawa, Piotr Syga

2021

Abstract

DeepFakes are videos that include maliciously added in the postprocessing changes, quite often substituting face of a portrayed individual with a different face using neural networks. Even though the technology gained its popularity as a carrier of jokes and parodies, it raises a serious threat to ones security – via biometric impersonation or besmearing. In this paper we focus on a method that allows of detecting DeepFakes for a user without significant computational power. In particular, we enhance MesoNet (Afchar et al., 2018) by replacing the original activation functions. We achieve over 1% improvement as well as increasing the consistency of the results. Moreover, we introduced and verified a new activation function — Pish that at the cost of slight time overhead allows even higher accuracy on certain datasets.

Paper Citation

in Harvard Style

Kawa P. and Syga P. (2021). Verify It Yourself: A Note on Activation Functions’ Influence on Fast DeepFake Detection. In Proceedings of the 18th International Conference on Security and Cryptography - Volume 1: SECRYPT, ISBN 978-989-758-524-1, pages 779-784. DOI: 10.5220/0010581707790784

in Bibtex Style

@conference{secrypt21,
author={Piotr Kawa and Piotr Syga},
title={Verify It Yourself: A Note on Activation Functions’ Influence on Fast DeepFake Detection},
booktitle={Proceedings of the 18th International Conference on Security and Cryptography - Volume 1: SECRYPT,},
year={2021},
pages={779-784},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0010581707790784},
isbn={978-989-758-524-1},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 18th International Conference on Security and Cryptography - Volume 1: SECRYPT,
TI - Verify It Yourself: A Note on Activation Functions’ Influence on Fast DeepFake Detection
SN - 978-989-758-524-1
AU - Kawa P.
AU - Syga P.
PY - 2021
SP - 779
EP - 784
DO - 10.5220/0010581707790784