DeepFake-o-meter: An Open Platform for DeepFake Detection

View on arXiv ← Back to list

Authors: Yuezun Li, Cong Zhang, Pu Sun, Honggang Qi, Siwei Lyu

Published: 2021-03-02 20:45:33+00:00

AI Summary

This paper introduces DeepFake-o-meter, an open-source online platform that integrates multiple state-of-the-art deepfake detection methods. The platform provides a user-friendly interface for detecting deepfakes in videos and aims to facilitate research and development in this field.

Abstract

In recent years, the advent of deep learning-based techniques and the significant reduction in the cost of computation resulted in the feasibility of creating realistic videos of human faces, commonly known as DeepFakes. The availability of open-source tools to create DeepFakes poses as a threat to the trustworthiness of the online media. In this work, we develop an open-source online platform, known as DeepFake-o-meter, that integrates state-of-the-art DeepFake detection methods and provide a convenient interface for the users. We describe the design and function of DeepFake-o-meter in this work.

Key findings

The DeepFake-o-meter platform successfully integrates over ten state-of-the-art deepfake detection methods. The platform provides a convenient interface for users and researchers to access and compare these methods. Future work will focus on expanding the platform's capabilities and improving its efficiency.

Approach

The platform integrates various existing deepfake detection models into a unified framework. It uses a client-server architecture with a web-based frontend for user interaction and a backend employing Docker containers to manage the diverse models and their dependencies. The results are visualized and made available to the users.

Datasets

UNKNOWN

Model(s)

MesoInception4, FWA, VA-MLP, Xception-c23, ClassNSeg, Capsule, DSP-FWA, CNNDetection, Upconv, WM, Selim

Author countries

China, China, USA

← Previous