A Lightweight and Interpretable Deepfakes Detection Framework

View on arXiv ← Back to list

Authors: Muhammad Umar Farooq, Ali Javed, Khalid Mahmood Malik, Muhammad Anas Raza

Published: 2025-01-21 07:03:11+00:00

AI Summary

This paper proposes a lightweight and interpretable deepfake detection framework that uses a fusion of facial landmark and novel heart rate features. These features are fed into an XGBoost classifier to distinguish between real and deepfake videos, achieving superior performance compared to existing methods while maintaining interpretability.

Abstract

The recent realistic creation and dissemination of so-called deepfakes poses a serious threat to social life, civil rest, and law. Celebrity defaming, election manipulation, and deepfakes as evidence in court of law are few potential consequences of deepfakes. The availability of open source trained models based on modern frameworks such as PyTorch or TensorFlow, video manipulations Apps such as FaceApp and REFACE, and economical computing infrastructure has easen the creation of deepfakes. Most of the existing detectors focus on detecting either face-swap, lip-sync, or puppet master deepfakes, but a unified framework to detect all three types of deepfakes is hardly explored. This paper presents a unified framework that exploits the power of proposed feature fusion of hybrid facial landmarks and our novel heart rate features for detection of all types of deepfakes. We propose novel heart rate features and fused them with the facial landmark features to better extract the facial artifacts of fake videos and natural variations available in the original videos. We used these features to train a light-weight XGBoost to classify between the deepfake and bonafide videos. We evaluated the performance of our framework on the world leaders dataset (WLDR) that contains all types of deepfakes. Experimental results illustrate that the proposed framework offers superior detection performance over the comparative deepfakes detection methods. Performance comparison of our framework against the LSTM-FCN, a candidate of deep learning model, shows that proposed model achieves similar results, however, it is more interpretable.

Key findings

The proposed framework outperforms existing methods on the WLDR dataset, achieving comparable accuracy to deep learning models like LSTM-FCN but with improved interpretability. The fusion of facial landmark and heart rate features significantly improves detection performance.

Approach

The framework extracts 850-D facial landmark features using OpenFace2 and 63-D novel heart rate features from seven facial regions. These features are fused and standardized before being input into an XGBoost classifier for deepfake detection.

Datasets

World Leaders Dataset (WLDR)

Model(s)

XGBoost

Author countries

Pakistan, USA

← Previous