In Ictu Oculi: Exposing AI Generated Fake Face Videos by Detecting Eye Blinking

TL;DR

This paper presents a novel method for detecting AI-generated fake face videos by analyzing eye blinking patterns, exploiting the fact that synthetic videos often lack realistic blinking, and demonstrates promising results on benchmark datasets.

Contribution

The paper introduces a new eye-blinking detection approach to identify deepfake videos, addressing a key limitation in current generative models.

Findings

Effective detection of fake videos using eye-blinking cues

High accuracy on benchmark eye-blinking datasets

Promising results on DeepFake videos

Abstract

The new developments in deep generative networks have significantly improve the quality and efficiency in generating realistically-looking fake face videos. In this work, we describe a new method to expose fake face videos generated with neural networks. Our method is based on detection of eye blinking in the videos, which is a physiological signal that is not well presented in the synthesized fake videos. Our method is tested over benchmarks of eye-blinking detection datasets and also show promising performance on detecting videos generated with DeepFake.