Deepfake Detection using Biological Features: A Survey

View on arXiv ← Back to list

Authors: Kundan Patil, Shrushti Kale, Jaivanti Dhokey, Abhishek Gulhane

Published: 2023-01-14 05:07:46+00:00

AI Summary

This survey paper reviews deepfake detection methods focusing on biological features like eyebrow recognition, eye blinking, eye movement, ear and mouth detection, and heartbeat detection. It compares different biological features and their classifiers, highlighting challenges and proposing future research directions.

Abstract

Deepfake is a deep learning-based technique that makes it easy to change or modify images and videos. In investigations and court, visual evidence is commonly employed, but these pieces of evidence may now be suspect due to technological advancements in deepfake. Deepfakes have been used to blackmail individuals, plan terrorist attacks, disseminate false information, defame individuals, and foment political turmoil. This study describes the history of deepfake, its development and detection, and the challenges based on physiological measurements such as eyebrow recognition, eye blinking detection, eye movement detection, ear and mouth detection, and heartbeat detection. The study also proposes a scope in this field and compares the different biological features and their classifiers. Deepfakes are created using the generative adversarial network (GANs) model, and were once easy to detect by humans due to visible artifacts. However, as technology has advanced, deepfakes have become highly indistinguishable from natural images, making it important to review detection methods.

Key findings

The survey reveals that biological features offer promising avenues for deepfake detection. However, the accuracy of these methods varies depending on the feature and classifier used, and high-quality deepfakes pose significant challenges. Future research should address the limitations and explore the impact of video processing and makeup on the effectiveness of these techniques.

Approach

The paper surveys existing deepfake detection techniques that utilize biological features. It analyzes the performance of different classifiers for each feature (eyebrow, eye blinking, eye movement, ear/mouth, heartbeat) and discusses their strengths and limitations.

Datasets

Celeb-DF, FaceForensics++, DFDC Preview, FakeET, Google/Jigsaw Deepfake Detection, HOHA dataset, CEW Dataset, UBIRIS, CASIA

Model(s)

CNN, RNN, LSTM, LRCN, LightCNN, ResNet, DenseNet, SqueezeNet, Xception net, MLP, SVM, VGG16, ResNet50, Dec-Tree, T-CNN, Naive Bayes, autoregressive (AR) model

Author countries

UNKNOWN

← Previous