Generalized Deepfakes Detection with Reconstructed-Blended Images and Multi-scale Feature Reconstruction Network

TL;DR

This paper introduces a universal deepfake detection method that uses reconstructed blended images and a multi-scale feature reconstruction network to effectively identify face manipulations across various datasets and unseen forgeries.

Contribution

It proposes a novel blended-based training sample generation and a multi-scale feature reconstruction network for robust deepfake detection, improving cross-dataset and cross-manipulation performance.

Findings

Enhanced detection accuracy on unseen datasets.

Effective identification of boundary artifacts and noise anomalies.

Robust performance across different manipulation techniques.

Abstract

The growing diversity of digital face manipulation techniques has led to an urgent need for a universal and robust detection technology to mitigate the risks posed by malicious forgeries. We present a blended-based detection approach that has robust applicability to unseen datasets. It combines a method for generating synthetic training samples, i.e., reconstructed blended images, that incorporate potential deepfake generator artifacts and a detection model, a multi-scale feature reconstruction network, for capturing the generic boundary artifacts and noise distribution anomalies brought about by digital face manipulations. Experiments demonstrated that this approach results in better performance in both cross-manipulation detection and cross-dataset detection on unseen data.