electronics

Review
A Contemporary Survey on Deepfake Detection: Datasets,
Algorithms, and Challenges
Liang Yu Gong *,† and Xue Jun Li *,†

Department of Electrical and Electronic Engineering, Auckland University of Technology,

Auckland 1010, New Zealand
* Correspondence: liangyu.gong@autuni.ac.nz (L.Y.G.); xuejun.li@aut.ac.nz (X.J.L.)
† These authors contributed equally to this work.

Abstract: Deepfakes are notorious for their unethical and malicious applications to achieve economic,
political, and social reputation goals. Recent years have seen widespread facial forgery, which does
not require technical skills. Since the development of generative adversarial networks (GANs) and
diffusion models (DMs), deepfake generation has been moving toward better quality. Therefore, it
is necessary to find an effective method to detect fake media. This contemporary survey provides
a comprehensive overview of several typical facial forgery detection methods proposed from 2019
to 2023. We also analyze and group them into four categories in terms of their feature extraction
methods and network architectures: traditional convolutional neural network (CNN)-based detection,
CNN backbone with semi-supervised detection, transformer-based detection, and biological signal
detection. Furthermore, it summarizes several representative deepfake detection datasets with their
advantages and disadvantages. Finally, we evaluate the performance of these detection models with
respect to different datasets by comparing their evaluating metrics. Across all experimental results
on these state-of-the-art detection models, we find that the accuracy is largely degraded if we utilize
cross-dataset evaluation. These results will provide a reference for further research to develop more
reliable detection algorithms.

Keywords: deepfake detection; deep learning methods; transformer; semi-supervised learning;

evaluating metrics; state-of-the-art models
Citation: Gong, L.Y.; Li, X.J. A
Contemporary Survey on Deepfake
Detection: Datasets, Algorithms, and
Challenges. Electronics 2024, 13, 585.
1. Introduction
https://doi.org/10.3390/
electronics13030585
Although deepfakes were initially associated with entertainment such as movie visual
effects, camera filters, and digital avatars [1], they are defined as “believable generated
Academic Editors: Chiman Kwan and media by Deep Neural Network” and have evolved into a mainstream tool for facial forgery.
Beiwen Li
Their illegal applications now pose serious threats to social stability, national security,
Received: 16 December 2023 and personal reputation [2]. Facial manipulation technologies started with 3D landmark
Revised: 23 January 2024 face swap and auto-encoders [3] to generate fake media; however, the trend of deepfake
Accepted: 29 January 2024 generation nowadays involves more powerful generative models such as generative and
Published: 31 January 2024 adversarial networks (GANs) [4,5] and diffusion models (DMs) [6] for creating more
realistic counterfeit media. As for the illegal application of this technique, one Reddit user
first released generated pornographic videos of actress Gal Gadot as the protagonist of
deepfakes, which caused a huge sensation and harmed the victim’s reputation at the end
Copyright: © 2024 by the authors. of 2017. In addition, Rana et al. [7] found that the top ten pornographic websites have
Licensee MDPI, Basel, Switzerland. released over 1790 deepfake videos to transfer celebrities’ faces to porn stars’ faces. To
This article is an open access article
address these threats, deepfake detection and its performance have attracted significant
distributed under the terms and
consideration both in the academic and industrial fields; for example, Facebook, Microsoft,
conditions of the Creative Commons
and Amazon jointly launched a Deep Fake Detection Challenge (DFDC) [8] on Kaggle
Attribution (CC BY) license (https://
from 2019 to 2020. Meanwhile, we compared the publications of deepfake generation
creativecommons.org/licenses/by/
with detection from Dimensions (See Figure 1), which is a scholarly database that goes
4.0/).

Electronics 2024, 13, 585. https://doi.org/10.3390/electronics13030585 https://www.mdpi.com/journal/electronics

Electronics 2024, 13, 585 2 of 22

beyond research articles and their citations; then, we surprisingly noticed a shift in the
publication trend, with deepfake detection surpassing deepfake generation in the past
two years. Therefore, it is necessary to conduct in-depth research on deepfake detection
methods for future investigators to study in order to prevent fraud through facial forgery
or illegal information dissemination via deepfakes.

(a)

(b)

Figure 1. Relative publication data obtained from Dimensions database [9] at the end of 2023 by
searching “deepfake generation” and “deepfake detection” as keywords: (a) number of deepfake-
generation-related scholarly papers from 2014 to 2023; (b) number of deepfake-detection-related
papers from 2014 to 2023.

Thus, a comprehensive literature review of deepfake detection will be useful for re-
searchers to study this field further in different aspects with the development of deepfake
generation. This motivates us to present a deepfake detection survey in review of (1) deep-
fake detection databases, (2) categorized several typical deepfake detection of frame-based
and video-based methods, (3) the latest trend of detection methods (biological signal based),
and (4) a summary and analysis of future trend of deepfake detection. Specifically, the
main contributions of this survey are twofold.
Electronics 2024, 13, 585 3 of 22

1. We summarize and categorize detection techniques such as self-consistency, consistency-

based detection, detection with vision transformer, contrastive learning for detection,
inconsistency detection, biological signal-based detection, and a specially designed
network (Capsule Network).
2. We perform an in-depth comparison in terms of evaluating the metrics of three state-of-
the-art fake detectors: (i) Capsule Network, (ii) Consistency Learning Representation
(CORE), and (iii) T-Face Detection Model.

2. Deepfake Detection Datasets

Most online face forgery tools (such as DeepFaceLive [10] and Roop [11]) are open
source and do not require sophisticated technical skills, so using open-source software
such as Basic DeepFake maker [12] is the main method for creating deepfake datasets.
Due to multiple forgery methods, deepfake data are increasing at a very high rate of
approximately 300% every year [2], but the data published online have different forgery
qualities. This section introduces several representative datasets and illustrates their
advantages and disadvantages.

2.1. FaceForensics++
FaceForensics++ [13] is a pioneering large-scale dataset in the field of face manipulation
detection. The main facial manipulations are representative, which include DeepFakes,
Face2Face, FaceSwap, FaceShifter, and Neural Textures methods, and data are of random
compression levels and sizes [14]. This database originates from YouTube videos with
1000 real videos and 4000 fake videos, the content of which contains 60% female videos
and 40% male videos. In addition, there are three resolutions of videos: 480p (VGA), 720p
(HD), and 1080p (FHD). As a pioneering dataset, it has different quality levels of data and
equalized gender distributions. The deepfake algorithms include face alignment and Gauss–
Newton optimization. However, this dataset suffers from low visual quality with high
compression and visible boundaries of the fake mask. The main limitation of this dataset is
the lack of advanced color-blending processing, resulting in some source facial colors being
easily distinguishable from target facial colors. In addition, some target samples cannot
effectively fit on the source faces because there exists facial landmark mismatch, which is
shown in Figure 2.

Figure 2. Several FaceForensics++ samples. The manipulated methods are DeepFakes (Row 1),
Face2Face (Row 2), FaceSwap (Row 3), and Neural Textures (Row 4). DeepFakes and FaceSwap
methods usually create low-quality manipulated facial sequences with color, landmark, and boundary
mismatch. Face2Face and Neural Textures methods can output slightly better-quality manipulated
sequences but with different resolutions.
Electronics 2024, 13, 585 4 of 22

2.2. DFDC
From 2020 to 2023, Facebook, Microsoft, Amazon, and research institutions put efforts
into this field and jointly launched a Deep Fake Detection Challenge (DFDC) [8] on Kaggle
to solve the problem of deepfakes presenting realistic AI-generated videos of people
performing illegal activities, with a strong impact on how people determine the legitimacy
of online information. The DFDC dataset is currently the largest public facial forgery
dataset, which contains 119,197 video clips of 10 s duration filmed by real actors. The
manipulation data (See Figure 3) are generated by deepfake, GAN-based, and non-learned
techniques with resolutions ranging from 320 × 240 to 3840 × 2160 and frame rates from
15 fps to 30 fps. Compared with FaceForensics++, this database has a large-enough sample
amount, different poses, and a rich diversity of human races. In addition, the original
videos are from 66 paid actors instead of YouTube videos, and fake videos are generated
with similar attributes to original real videos. However, the main drawback is that the
quality level of data is different due to several deepfake generative abilities. Therefore,
some samples have the problem of boundary mismatch, and source faces and target faces
have different resolutions.

Figure 3. DFDC samples. We manually utilized InsightFace facial detection model to extract human
faces from the DFDC. Although some of the samples are without color blending and with obvious
facial boundaries, the average quality is a little higher than the first-generation deepfake datasets.

2.3. Celeb-DF V2
Celeb-DF V2 is derived from 590 original YouTube celebrity videos and 5639 manip-
ulated videos generated through FaceSwap [15] and DFaker as mainstream techniques.
It consists of multiple age, race, and sex distributions with many visual improvements,
making fake videos almost indistinguishable to the human eye [16]. The dataset exhibits a
large variation of face sizes, orientations, and backgrounds. In addition, some post-processing
work is added by increasing the high resolution of facial regions, applying color transfer
algorithms and inaccurate face masks. However, the main limitation of this dataset is the
low data amount with less sample diversity because all original samples are downloaded
from YouTube celebrity videos, and there is small ethnic diversity, especially for Asian
faces.Here, we present a few samples of Celeb-DF V2 (see Figure 4).
Electronics 2024, 13, 585 5 of 22

Figure 4. Celeb-DF V2 crop manipulated facial frames. Except for transgender and transracial fake
samples (Row 3), it is hard to distinguish real and fake images with the human eye.

There are other higher-quality deepfake datasets created by extensive application of

the GAN-based method; for example, DFFD [17], which was published in 2020, created
an entire synthesis of faces by StyleGAN [18]. Comparing datasets published after 2020
with previous datasets, it can be observed the data amount is much larger with multiple
forgery methods such as GAN and forgery tools. In addition, the original data sources are
not limited to online videos such as YouTube and also consist of videos shot by real actors.
Thus, we predict the trend of future DeepFakes datasets to be larger scale with various
forgery methods, multiple shooting scenarios, and different human races. We summarize
the advantages and disadvantages of several commonly used datasets in Table 1.

Table 1. The typical and commonly used datasets of facial forgery detection.

Datasets Real/Fake Data Source Methods Advantages Limitations

UADF (2018) [19] 49/49 YouTube FakeApp Early release Low data amount
Visible
FaceForensics++ FS, F2F, NT,
1000/5000 YouTube Multiple methods manipulated
(2019) DeepFakes, and FS
artifacts
DeepFake- Relatively good
363/3068 Actors DeepFakes Low data amount
Detection (2019) effects
Improved Realistic Less forgery
Celeb-DF (2019) 590/5639 YouTube
DeepFakes manipulation methods
DeepFakes and Different quality
DFDC (2020) 19,197/1,000,000 Actors Various techniques
GAN levels
DeeperForensics- Large-scale with Less forgery
50,000/10,000 Actors DeepFakes
1.0 (2020) different attributes methods
iFakeFaceDB (2020) 494,414/33,000 Previous dataset GAN Multi-scenarios Unknown
YouTube and GAN, DeepFakes, Large-scale and Different quality
DFFD (2020) 58,703/240,336
previous datasets and FakeAPP multi-techniques levels
DeepFaceLab, Multi-face and
FFIW10K (2021) 12,000/10,000 YouTube Unknown
FSGAN, and FS scenarios
FS: FaceSwap; F2F: Face2Face; NT: Neural Textures; FS: FaceShifter.

3. Deepfake Detection
Different survey articles have categorized deepfake detection methods from different
perspectives. For example, Rana et al. [7] group detection methods into four categories:
deep-learning-based techniques, classical machine-learning-based techniques, statistical
Electronics 2024, 13, 585 6 of 22

techniques, and blockchain-based techniques. Nguyen et al. [20] directly categorized

based on data type, i.e., video detection and image detection. These categorizations are
relatively general and not beneficial for a deeper understanding of the latest deepfake
detection algorithms and model architectures. In this section, we start from the perspective
of data augmentation, feature extraction, and loss function and illustrate their unique
model architectures to provide a new classification of deepfake detection methods. This
review is a summary of some of the innovative algorithms at the current stage. The aim of
this review paper is to study how a model can learn more manipulated information features.
Thus, it analyzes more data augmentation methods, feature extraction backbones, and loss
function designs, which will hopefully inspire researchers to study this field further.

3.1. Traditional CNN-Based Detection Methods

Traditional CNN-based methods [21] are widely applied to detect tampered images
by frequency-domain features and statistical features. The CNN architecture can increase
the learning capacity of a model by constantly stacking convolutional blocks and shortcut
connections. Previous research has already proven that several models, such as Xception-
Net and ResNet, can effectively extract fake features caused by compression and device
fingerprints. Thus, we can find that most newly published deepfake detection methods
choose these networks as the backbone for feature extraction. With the development
of forgery techniques, manipulated media always come with advanced post-processing,
which shows that deepfakes can be more realistic, and a detailed forgery clue is almost
hard to distinguish from real ones. Thus, the similarity distances between extracted
deepfake and real features are shorter than previously, and it is harder for a model to
distinguish these samples by binary classification algorithms such as the support vector
machine (SVM) because the extracted features are too concentrated on a specific area of the
hyperplane. This is the biggest limitation of traditional CNN-based models for frame-based
deepfake detection because the inconsistency between the frames is not considered at
this stage. Usually, the unseen data come with different manipulation qualities, and the
traditional CNN-based method cannot overperform on unlearned features. Thus, this
category cannot provide good results in the latest manipulation and can easily overfit. In
this section, we only focus on Capsule Network, which is one specially designed network
for deepfake detection.

Capsule Network
Capsule Network [22] itself was not a new term in the field of deep learning, and it
was first proposed to solve the problem that traditional convolutional neural networks
(CNNs) have limited applicability in “inverse graphics”. For classification work, traditional
CNNs aimed to stack convolutional layers to extract multi-scale features corresponding
to different receptive fields, where there is less consideration of the relationship between
different feature information. However, Capsule Network (See Figure 5) has one typical
strength: they are able to learn 3D spatial information about objects with their relationship
and then model them explicitly. Moreover, Capsule Network [23] can use fewer parameters
and data to perform similarly to CNN.
Specifically, Nguyen et al. [23] proposed a VGG-19 [24] feature extraction backbone,
which is from the first layer to the third layer as the backbone and pre-trained in ILSVRC
datasets [25], able to reduce overfitting and transfer learning. The input images are set to
300 × 300. The output features are passed through the backbone and sent to 10 primary
Capsule Networks and output 4 × 1 vector capsules by a dynamic routing algorithm. Each
primary Capsule Network [26] consists of two 2D CBL modules (convolution layer + batch
normalization + ReLU), a statistical pooling layer and two 1D CB modules (1D convolution
layer + batch normalization), and the overall CapsuleNet includes the parallel connection
of ten primary Capsule Nets (see Figure 6). The parameters of convolutional blocks are
shown in Figure 6 as well. Another difference between CapsuleNet and CNN is the output.
Dynamic routing is designed to calculate how the output of lower-level capsules is allocated
Electronics 2024, 13, 585 7 of 22

to higher-level capsules, which is achieved by the squash activation function shown in

Equation (1):
∥ u ∥2 u
squash(u) = 2 ∥u∥
(1)
1 + ∥u∥
where the left side is the scale factor obtained by the extracted features, and the right side
is the unit vector.

Figure 5. The architecture of CapsuleNet faces forensics detection. This method utilizes the CNN
backbone to extract features and Capsule Network to output vectors for prediction.

Figure 6. Details of the primary Capsule Network structure with relative parameters. Each Capsule
Network includes the parallel connection of 10 primary capsules.

In summary, Capsule Network can study 3D spatial information relationships, which

is better than the traditional CNN model, and it utilizes fewer parameters. However,
its drawback is the weak generalization capacity that Capsule Network cannot perfectly
predict unseen data. A detailed explanation is provided in Section 3.5.2.

3.2. CNN Backbone with Semi-Supervised Learning

The semi-supervised method is a model learning method that is between supervised
learning and unsupervised learning, which means that the final decision is not only de-
termined by the projection of features and labels. Compared with supervised learning
methods, it requires certain data augmentations to enrich current datasets to data pairs and
further similarity calculation on data. Most semi-supervised learning methods are achieved
by three approaches, namely, context-based learning, temporal-based learning, and con-
trastive learning, and these approaches are widely applied in frame-based detection and
video-based detection. In this section, we will mainly investigate four representative detec-
tion methods: temporal detection, consistency learning, inconsistency learning detection,
and contrastive learning detection.
Electronics 2024, 13, 585 8 of 22

3.2.1. Consistency Representation Learning

COnsistency REpresentation Learning of Forgery Detection (CORE) [27] is an effective
network that can capture different representations between two data augmentations and
regularize feature similarity by the cosine distance to enhance consistency so that it can
achieve relatively good results in both in-dataset and cross-dataset evaluation. We acknowl-
edge that traditional classifiers always follow the steps of data augmentation, extracting
and encoding features, and fully connected layers to achieve classification tasks. However,
CORE presents a new architecture (see Figure 7) focusing on the consistency of sample
pairs generated by different data augmentation strategies such as random resized crop and
random erasing (RE) [28], for example. RE augmentation utilizes a scale factor from 0.02 to
0.2 and an aspect ratio from 0.5 to 2 to cut out a region in the face. For each input image,
there is only a 0.33 probability without data augmentation, which ensures the input images
are various. CORE proposes a shared parameter Xception encoder [29] that separately
extracts features of data pairs, maps them into two-dimensional representation vectors in
the latent space, and then calculates their pixel-wise consistency loss. In addition, cosine
consistency loss can avoid forcing two representations to be totally the same. Finally, two
representations obtained from the backbone are fed to two classifiers trained by cross-
entropy loss as well and penalize the distances between two features adopted by cosine
loss. The total loss is guided by the linear combination of cosine loss and cross-entropy
loss, which is calculated by
Lcos = (1 − f 1n · f 2n )2 (2)
N
Lc = ∑ Lcos (3)
1

Lloss = LCE + αLc (4)

where f 1n
and f 2n
represent the normalization of Feature 1 and Feature 2, respectively; LCE
represents the cross-entropy loss; α represents the balance weight for cross-entropy loss
and consistency loss; and N represents the pairs of data augmentation.

Figure 7. The architecture of CORE. One method extracts pairs of representations by data augmenta-
tion and calculates consistency loss to guide final loss function.

Thus, CORE is used to observe the similarity between two views’ representations
obtained from data augmentations and feature extraction because we regard that each
type of data should be consistent even though applying different data augmentations. The
main strength of this method is to utilize a new loss function that the classification work is
determined by both cross-entropy loss and representation similarity after different data
Electronics 2024, 13, 585 9 of 22

augmentations. It largely increases the generalization ability of previous detection methods

because it shows better evaluating metrics on cross-dataset experiments. However, the
performance of this method largely relies on the data augmentation stage; thus, different
augmentations will influence the final evaluating metrics.

3.2.2. Self-Consistency Method

Most defense methods focus on detecting suspicious artifacts of fake media such as
eye blinking, blending boundaries [30], and face warping [31]. However, the manipulated
clues can be observed in the feature patches as well. Zhao et al. [32] proposed a method
for finding out the inconsistency of content-independent and spatially local information
within manipulated images by designing a multi-task learning architecture, as shown in
Figure 8, including the consistency branch and classification branch. More specifically, it
regards the manipulated images to contain different source features, which come from the
imaging pipeline (photo-response non-uniformity [33]) and encoding approaches (JPEG
compression pattern [34]) at different locations. Thus, this is a detection method mainly
based on calculating the similarity between different source feature patterns by pairwise
self-consistency learning similarity score in the consistency branch. The backbone selects
ResNet-34 [35] initialized by pre-trained weights on ImageNet for extracting pre-processed
video frames’ features. Then, the feature map (H × W × C) is divided into several patches,
and each patch is compared against the rest of the patches to compute dot product similarity
(consistency score):
θ ( f i )θ ( f j )
s( f i , f j ) = δ( √ ) (5)
C
where f i and f j represent patches of the feature, δ represents the sigmoid function, C
represent channel size, and θ represents the embedding function.

Figure 8. The architecture of the consistency branch and classification branch [32], One method
calculates the patch similarity and classification loss to guide the final loss.

By iteration of this process, finally, a 4D consistency volume will be obtained, and

ground-truth 4D volume is created by bi-linear downsampling and computation of
element-wise differences, where 1 denotes patches are consistent and 0 otherwise. In
addition, the inconsistency image generator utilizes elastic deformation to improve various
masks so it can eliminate spurious correlations. Then, it randomly selects the color-blending
method to improve feature robustness as well as data augmentation such as JPEG compres-
sion, Gaussian noise, and color jittering. The pairwise consistency learning (PCL) loss uses
binary cross-entropy loss (BCE Loss) to supervise:

1
L PCL =
N ∑ BCE(Vpred , VGT ) (6)
Electronics 2024, 13, 585 10 of 22

In short, this method also utilizes two factors to determine the final loss function.
However, the difference is to find the patches’ similarity and not the consistency between
different representations because it is based on the notion that the forgery clues of fake
video generation will project to different feature patches. In addition, it requires performing
ablation experiments on four datasets, including DFR, CD2, DFDC, and DFDC-P, and the
average area under the curve (AUC) reached 82% above, which is a significant development
in forgery detection. However, there are two limitations: this method cannot detect entire
facial synthesis created by GAN or diffusion models, and it can be further improved on
low-quality data.

3.2.3. Dual-Contrastive Learning Detection (T-Face)

Recently, Youtu Lab [36] proposed a dual-contrastive learning architecture that is
aimed at distinguishing the authenticity of faces by different data augmentation. We found
out that traditional detection methods like CNN-based methods are not suitable for de-
tecting and learning generalized feature representations. Dual-contrastive learning (DCL)
includes inter-instance contrastive learning and intra-instance contrastive learning for
narrowing the distance between embeddings of the same class and widening the distances
between non-homogeneous embeddings. Unlike traditional data augmentation, T-face
consists of a data views generation (DVG) module with four data enhancements to elimi-
nate task-irrelevant contextual factors—random patch, high-frequency enhancement [37],
frame shift, and corresponding mix-up, which are selected with certain probability—and
randomly combine with traditional data augmentation to generate two data views, V1 and
V2 . The details of four data augmentation factors are shown in Table 2.

Table 2. Details of four data augmentation factors.

Data Augmentation Details Purposes

Divide facial area into k × k patches and
Random patch Destroy facial structures
randomly shuffle and reassemble them
Features through SRM model and combine
High-frequency enhancement Boost generalization ability
with original image
Frame shift Select multiple frames from one video Reduce the influences of motions and expressions
Associate fake images with corresponding
Corresponding mix-up Eliminate some obvious forgery clues
real images

The architecture of DCL (see Figure 9) includes two branches of feature extraction
encoders ( f q and f k); then, the batch images are fed to encoders to obtain extracted features
by applying 1 × 1 convolution operation to squeeze the channel to obtain the query q
and the key k separately. The encoders utilize the strategy of the exponential moving
average. The parameters of the key encoder are updated from query encoder parameters.
It promotes generalized feature learning by cross-entropy loss, InfoNCE loss [38], and
intra-instance contrastive loss together. For the performance of T-face, Sun et al. [36]
conducted a cross-dataset experiment on FF++ (trained), DFD, DFDC, Wild DeepFake, and
Celeb-DF. The AUC metric and equal error rate (EER) presented relatively good results on
Celeb-DF and DFD, which are shown in Table 3.

Table 3. Experiments of DCL [36] in different datasets.

Datasets AUC EER

FF++ 99.30% 3.26%
DFD 91.66% 16.63%
DFDC 76.71% 31.97%
Wild Deepfake 71.14% 36.17%
Celeb-DF 82.30% 26.53%
Electronics 2024, 13, 585 11 of 22

Figure 9. An overview of DCL architecture. Reprinted with permission from Ref. [36]. 2024, K.Sun.
Four random data augmentation factors were utilized in this contrastive learning method to guide
inter-class and intra-class distances separately.

3.2.4. Spatio-Temporal Inconsistency Learning

Video-based detection such as LSTM [39] cannot achieve better results than the above
frame-based detection methods in the current stage, but learning temporal artifacts has
already been proven to increase detection robustness. Thus, Gu et al. presented a detection
work [40] based on learning fake videos’ spatial and temporal inconsistency. To our knowl-
edge, it is quite difficult to distinguish one frame as real or fake because of the developed
manipulation techniques; however, if we combine frames and observe their temporal in-
consistency such as face position jittering, detection will be easier. A novel STIL block (see
Figure 10) works in a two-stream manner and contains a temporal inconsistency module
(TIM), a spatial inconsistency module (SIM), and an information supplement module (ISM),
which are proposed to obtain a representation with the spatial and temporal information.

Figure 10. The structure of STIL [40]. Each STIL block contains SIM, TIM, and ISM modules.

Firstly, the input sequence [ T, C, H, W ] was split along the channel dimension into two
portions, where each portion is a feature with a size of [ T, C2 , H, W ]. Then, two portions were
separately fed as the inputs of TIM and SIM to acquire two inconsistency representations
from spatial and temporal perspectives. Specifically, SIM and TIM were designed as three
branch modules, as shown in Tables 4 and 5, which aimed to find pixel-wise boundary
mismatch and temporal inconsistency. Based on the ablation experiments, STIL proposed
that the best performance occurs when we fuse SIM modules into TIM modules.
TIM modules utilize temporal difference calculation, which is the subtraction of
adjacent frame features.
sth = Conv1( xth+1 − xth ) (7)
where Conv1 is 3 × 1 convolutional layer, and xth+1 and xth are features of t + 1 times of
frame and t times of frame, respectively.
Electronics 2024, 13, 585 12 of 22

Table 4. SIM branches.

Branches Operations
Upper branch ResNet block for shortcut connection
Downsampling, utilize 1 × 3 and 3 × 1 convolutional layer to obtain
Middle branch
vertical and horizontal features and upsampling
Confidence calculation Fuse the upper and middle features to obtain confidence by sigmoid
Bottom branch 3 × 3 convolutional layer and multiply with confidence

Table 5. TIM branches.

Branches Operations
Convolution and reshape, temporal difference calculation, and vertical
Upper branch
temporal inconsistency enhancement
Convolution and reshape, temporal difference calculation, and horizontal
Middle branch
temporal inconsistency enhancement
Bottom branch ResNet block for shortcut connection

3.3. Transformer-Based Detection

Although the transformer network was first designed for learning long contextual
sequence information and solving natural language tasks such as machine translation,
Dosovitskiy et al. [41] showed that it is not necessary to highly rely on CNN to perform
image classification tasks, and a pure transformer applied to image patch embeddings can
perform better on classification as well. Vision transformer (VIT) divides an image into
several flattened patches and embeds them into patch sequences first; then, it applies an at-
tention mechanism to obtain each patch’s attention weight and extract features. This means
that the transformer framework on classification is gradually replacing traditional CNN
classification even though there is limited exposure to deepfake detection. For example,
there is a video transformer with the incremental learning approach [42] published in 2021,
which utilizes Xception [29] as the backbone and 12 transformer blocks to learn feature
correlations. In addition, most studies are still using CNN as a feature extractor and trans-
former block to classify deepfake images such as the convolutional vision transformer [43].
In this section, we introduce two transformer-based methods, which are the end-to-end
transformer architecture and the video transformer.

3.3.1. End-to-End Transformer Detection

Most transformer-based deepfake detection only utilizes transformer blocks in
downstream tasks, such as studying the correlation relationship after extracting fea-
tures by the CNN-based backbone. However, the community of vision transformer clas-
sification lacks an end-to-end transformer detection framework for deepfake detection.
DFDT [44] proposes an end-to-end deepfake detection framework (see Figure 11) using
vision transformers that can basically solve the problem that traditional CNN cannot obtain
the correlation between spatial patches and the information loss caused by the receptive
field. It comprises four basic components: patch embedding [45], multi-stream transformer
block, attention-based patch selection, and multi-scale classifier. After the pre-processing of
videos, each video selects 20% video frames as the input to focus on facial regions only. In
addition, the two-stream transformer block is designed to detect forgery clues in different
facial regions by dividing input images into different patch sizes. The low-level patch
branch processes smaller facial regions, such as lips, and the high-level patch branch is
for extracting larger facial manipulated features, such as boundary mismatch. Then, the
patch embedding is different from vision transformer’s embedding because it realizes that
non-overlapping feature extraction, where the sliding windows’ stride equals the patch
size, will harm the neighboring local structures. In other words, this method chooses
overlapping patch extraction where two adjacent patches share a specific area, and it will
preserve neighboring information better. Similarly, with a class token added, the patches
Electronics 2024, 13, 585 13 of 22

flatten the positional embedding and project it into a latent space. A patch selection mech-
anism based on attention weight is applied in the second module, which can pay more
attention to sensitive information and dismiss the less useful information in the training
phase. Once the two-stream transformer blocks output the feature with the patch attention
weight, they will both make initial predictions. The average of all the predicted results in
the final decision.

Figure 11. An overview of DFDT framework for deepfake detection [44]. It includes overlapping
patch embedding, patch selection mechanism, multi-scale transformer block, and classifier.

In summary, this detection method proposes an end-to-end vision transformer frame-

work instead of utilizing CNN to extract features and choosing VIT to perform downstream
classification. DFDT can capture different scales of alterations and achieve SOTA results
both in in-dataset experiments and cross-dataset experiments, which means it has better
generalization ability and strong effectiveness, and the transformer can be a better feature
extractor than most CNN backbones.

3.3.2. ISTVT: A Video Transformer for Deepfake Detection

Although vision transformers have made significant progress in several vision clas-
sification tasks, research on video transformers is scarce. In detail, the video transformer
is a video-based detection method, which means it processes multiple video frames at
the same time and applies the self-attention mechanism on different token dimensions.
Following this inspiration, Zhao et al. [46] considered a video transformer that can jointly
study spatial and temporal information in fake videos and has better performance and
generalization in the deepfake detection field. ISTVT (see Figure 12) consists of a spatial–
temporal self-attention module and a self-subtract module to capture spatial features and
temporal inconsistency in videos. Specifically, the proposed video transformer first utilizes
Xception as a backbone to extract features in video sequences; then, it splits these features
by flattening them to obtain spatial, temporal, and class tokens, which can be fed to corre-
sponding transformer blocks as input. The tokens require additional pre-processing named
the self-subtract mechanism [46] to reduce the adjacent tokens in temporal dimensions.
It is used for generating feature residuals to ignore redundant information and focus on
temporal inconsistency. Similar to the vision transformer, the tokens are projected to Q, K,
and V by a linear projection layer with shape ( T + 1) × ( HW + 1) × C, and a self-attention
mechanism is applied both in the spatial dimension and the temporal dimension. A tempo-
ral attention block is inserted after each spatial attention block, and the relative position
Electronics 2024, 13, 585 14 of 22

embedding [47] applied on each temporal attention block is used to distinguish the order
of frames. Finally, the MLP block is selected for the final classification task.

Figure 12. An architecture of ISTVT [46]. It consists of four basic components: backbone feature
extraction, token embedding, self-attention blocks, and MLP.

The performance of cross-dataset experiments was tested, and the accuracy rates
of Celeb-DF, DFDC, and FaceShift were 84.1%, 74.2%, and 99.3%, respectively. With
the increased accuracy results, it is proved that the transformer with spatial–temporal
inconsistency detection reached higher generalization in unseen data than previous video-
based detection methods.

3.4. Biological Signal Detection

Although the above detection methods are relatively sophisticated and commonly
applied in the deepfake detection field, analyzing forgery artifacts in biological signals has
become a new trend that has attracted much attention. These methods focus on capturing
fake information by combining biological signals with vision artifacts that can achieve
better results or multi-modality detection. This section will only briefly introduce two
methods based on PPG cells and physiological signals. These methods are still developing
with good potential but always require specific matching biological maps such as heart rate.
Umur et al. [48] proposed an approach to extract PPG cells from both real and
deepfake videos and designed a classification network to distinguish authenticity and
related forgery methods. Specifically, this method proposed to first extract raw signals from
real videos and fake videos from different facial locations by windows, and then encode
these signals along with spectral density into a spatio-temporal block to create PPG cells.
Thus, the model input is no longer a batch of video frames instead of PPG cells, which
includes the projection of biological signals. Comparing the data pre-processing method
with other video-based detection methods, it is reasonable to believe that the residuals
learned by CNN with PPG biological information will obtain a better benchmark result,
and the spatial and temporal patterns of biological signals can be conceived and projected
to residuals [48]. The experimental accuracy finally reached 97.29% in detecting fake videos
and 93.39% in their related generative models by setting a window length of 64. Moreover,
remote photoplethysmography (rPPG) can extract heartbeat signals from recorded videos.
Wu et al. [49] realized that facial manipulation progress is inevitable with sudden facial
color changes in some periods, and rPPG could become an indicator for deepfake detection.
Therefore, it is necessary to obtain a multi-scale PPG map for classification, and the accuracy
of this method can reach more than 99% in the FF++ in-dataset experiment, which is a large
increase compared to Capsule Net and other CNN-based deepfake detectors.
In summary, the biggest contribution of this method is that it leads deepfake detection
toward source detection by adding biological signals to classify the residuals of generative
models. It also provides comparison experiments to illustrate that the accuracy will increase
by 47% by adding PPG cells with the same VGG-19 backbone, even outperforming the
Xception Network by more than 10%. These significant experimental results prove that
there is potential to add PPG signals to enhance the diversity of deepfake features with
better classification abilities. However, the selection of window size will influence the
Electronics 2024, 13, 585 15 of 22

stability and representative power of PPG cells because a small window will miss PPG
frequencies, and a too-large window will include more noise.
Physiological signals can be utilized to classify deepfake videos, and they can also
combine multi-modal information; for example, Stefanov et al. [50] proposed a method (See
Figure 13) to extract physiological signals and utilized the graph convolutional network
(GCN) [51] to fuse video and physiological signal and detect the dissimilarity between
audio and video modality. In particular, this method proposes an intriguing algorithm
to obtain visual physiological signals according to the following steps: first, facial areas
are detected with alignment, and background areas are removed; then, they are passed
through MTTS-CAN [52] and a square occlusion patch to estimate heart rate and respiration
rate, respectively. Finally, the difference between the estimated signal with occlusion
and without occlusion is considered as the relative contribution of the mask-out region
for the two physiological signals. In addition, one graph-based model is designed to
fuse the facial information with a visual physiological map. This information is used as
nodes with previously extracted features by ResNet18. Each feature is connected with the
physiological map and calculates the cosine similarity. By training the GCN model and
ResNet18, the model can combine two modalities and generate visual representations with
physiological signals.

Figure 13. Two approaches combining visual representations with physiological signals Reprinted
with permission from Ref. [50], 2024, Stefanov, K.

3.5. Experiment Procedures

3.5.1. Evaluating Metric Explanations
Accuracy is one standard used to evaluate metrics to illustrate model performance. How-
ever, there are two shortcomings when evaluating models. The first one is that it is not
reliable when positive and negative samples are unbalanced. Another drawback is that
it cannot present how many positive or negative samples are truly or wrongly predicted.
Thus, it is necessary to calculate the confusion matrix to obtain the true-positive rate (TPR),
true-negative rate (TNR), false-positive rate (FPR), and false-negative rate (FNR). Once
these four indicators are calculated, the AUC and EER are two metrics that can be used for
model evaluation.
There are three main evaluation metrics mentioned in this section, which are accuracy,
AUC, and EER. AUC and EER are calculated by the sklearn confusion matrix directly
in our experiments. Accuracy is the first evaluating metric that is commonly used in all
classification tasks. It is defined as the correctness of the prediction of positive and negative
samples. In this experiment, we regard one correct match between the predicted value
and the label value as one count. Accuracy is the ratio of count numbers to the total
sample number.
From previous analysis of evaluating metrics, there are four additional indicators
that illustrate model prediction performance, which are TPR, TNR, FNR, and FPR for
binary classification. These indicators can provide the greatest indication of how many
Electronics 2024, 13, 585 16 of 22

positive and negative samples are truly or falsely predicted. The confidence score of the
current sample is used as the threshold by traversing all samples. Through multiple sets
of thresholds, true-positive rate and false-positive rate pairs are calculated separately to
draw a curve, namely, the ROC curve. Thus, the area under this curve is the area under
the curve (AUC). Once the ROC is drawn, the equal error rate can be defined as the point
where the false-negative rate equals the false-positive rate. In conclusion, the larger the
accuracy and AUC values, the better the classification performance of the model. The
smaller the EER value, the better the model performs.

3.5.2. Cross-Dataset Applications

The experiments of this review paper utilize cross-dataset validation. The purpose
of cross-validation is to try to use different testing sets on the model to deal with too-
one-sided results and insufficient training data. It plays a key role in building machine
learning detection models because this process is not limited to the training phases and
evaluates a specific model’s predictive performance. Due to different forgery methods,
the model aimed to detect different kinds of manipulated samples effectively; thus, the
motivation is to evaluate model generalization ability and robustness when they have
already been perfectly fitted in one training set. If the evaluating metric drops significantly,
it proves that the generalization ability does perform well in the current forgery-validating
dataset; otherwise, the model is general and robust. However, most detection models
cannot effectively predict unseen forgery techniques at the feature level. It is reasonable
to observe the evaluating metrics such as accuracy drop slightly in cross-validation; on
the contrary, a large reduction in the evaluating metric is not acceptable, which means the
detector is useless.

3.5.3. Capsule Net

This experiment only used the FaceForensics++ [13] database as the training dataset.
On average, it selected 10 video frames of each video and obtained facial regions detected
by “scrfd kps.onnx”; finally, the training, validating, and testing cropped facial image
values are 15,783, 5259, and 5330, respectively, which are strictly followed by the ratio of
3:1:1. The Celeb-DF V2 [12] database followed the same operations, obtaining training,
validating, and testing data values of 24827, 8275, and 8278, respectively. For a performance
comparison, we performed multi-class training and binary class training separately and
then utilized the trained model to perform in-dataset and cross-dataset experiments with a
batch size of 16 accordingly. It firstly trained and tested Capsule Net only on multi-classes
(Face2Face, DeepFakes, FaceSwap, and Real video) of FaceForensics++ obtaining a testing
accuracy ranging from 92.2% to 98.73% in 30 epochs. Then, the binary class (Real or Fake)
experiments of FaceForensics++ obtained the best performance accuracy of 93.95% and an
EER of 4.34%, which shows good performance. However, when we utilized this model on
testing Celeb-DF V2, the evaluating metric decreased largely on the deepfake class, but
the original Real video class still displayed a relatively good result with a slight decrease.
To observe the testing performances on each class of the forgery method, we performed a
single-image prediction test on each class of FaceForensics++ and Celeb-DF via Capsule
Network, and the results are shown in Table 6.

Table 6. Experiments of Capsule Networks trained on FF++ only.

Original DeepFakes Face2Face FaceSwap Binary Test Binary Test

Capsule Net
ACC ACC ACC ACC ACC EER
Face Forensics++ 96.56% 98.58% 96.68% 94.50% 93.95% 4.34%
Celeb-DF V2 91.67% 28.87% 30.50% 34.50%
Electronics 2024, 13, 585 17 of 22

This experiment result proves that the testing accuracy of deepfakes is quite dependent
on the training data themselves by utilizing Capsule Network, and the performance indica-
tors of the forgery class are often greatly reduced in cross-dataset experiments, even though
we activated the dropout layer with a rate of 0.5. This is because the quality of databases is
different from each other, and we conducted an experiment combining some Celeb samples
into the training datasets. The accuracy of the testing results increased a lot, reaching 90%
above as well. Thus, we can make two bold assumptions that (1) the network formed by
stacking CBL modules and without any advanced data argumentation is not very effective
in extracting the features of all kinds of manipulated regions, and (2) the loss that is only
under the guidance of cross-entropy cannot provide reliable classification results.

3.5.4. CORE
The experiment utilized FF++ as the training set, which includes DeepFakes, Face2Face,
FaceSwap, Neural Textures, and one extra class of deepfake detection. Each video split
30 frames with the relative FF++ facial masks to obtain human facial regions, and the
training set and testing set strictly followed the FF++ data split requirements, which
had 86,263 fake images for the training phase and 16,641 fake images for the test. The
experiments explored and set a balance weight and cosine consistency loss to present the
benchmark area under the curve (AUC), which reached 99.96% on the FaceForensics++
in-dataset test in 30 epochs. The cross-dataset test results reached 72.41% and 75.72% on
DFDC and Celeb-DF, respectively. These results were also confirmed by our experiments.
From a previous literature review of the consistency representation method, this
method supposes the generalization ability of one detection model depends on the con-
sistency of its predicted results; that is, the real samples can correctly be predicted as real
samples, no matter the data augmented. Thus, it is necessary to complete an ablation
experiment to prove that the consistency of model prediction can be guaranteed to the
largest extent by a specific data augmentation method. As above, we acknowledged that dif-
ferent augmentation methods and balance weights will present different evaluating results.
Thus, we tested the ablation test on data augmentation, including RaAug, DFDC-Selim,
and Random Erasing, finding that the best performance of CORE’s data argumentation is
DFDC-Selim [53]. The reproduced test results are shown in Table 7.

Table 7. Ablation AUC results via different data argumentation.

Data Augmentation Celeb-DF V2

Random Erasing 74.78%
RaAug 67.83%
DFDC-Selim 79.45%

From the cross-dataset experiment on Celeb-DF, this detection method indeed largely
enhances the ability of generalization, but the performance is largely determined by data
augmentation methods and cannot reach a more reliable evaluating metric.

3.5.5. Spatio-Temporal Inconsistency Test

Although this experiment is a video-based detection method, it still requires the input
of several video frame sequences. We utilized “scrfd 10 g kps.onnx” to extract facial regions
with a crop area of 224 × 224. Then, we selected 8 frames to train this video-based model
and 16 frames to test. The batch size was set to 16, and we chose the Adam optimizer
to train the STIL model on one GPU within 30 epochs. After reproducing this project
with the following parameters, we calculated the in-dataset binary accuracy results of
Face Forensics++, Celeb-DF, DFDC, and ForgeryNet, as shown in Table 8. The cross-
dataset experiment on Celeb-DF, DFDC, and Wild-DF presented evaluating metrics of total
accuracy reaching 99.78%, 89.80%, and 84.12%, respectively.
Electronics 2024, 13, 585 18 of 22

Table 8. In-dataset testing accuracy results.

FF++ Celeb-DF DFDC ForgeryNet

Datasets
Real Fakes Real Fakes Real Fakes Real Fakes
Sample Numbers 140 560 125 790 1034 1500 1474 1836
Accuracy 94.29% 96.25% 90.04% 99.87% 91.20% 95.87% 98.64% 98.37%

3.5.6. Wrong Predicted Samples of Single-Image Prediction Test

Although most deepfake generation algorithms include post-processing blocks such
as color blending, the main challenge of the forgery algorithm is a clear forgery clue when
facial occlusion happens. To our knowledge, we believe these samples are not hard samples,
and the model should have a strong distinguishing ability for them. However, we provided
several manipulated samples (see Figure 14) generated by FaceSwap and Roop for a better
understanding of another detection challenge. We utilized two image detection methods,
which are Capsule Network and CORE, to perform the single-image prediction test, and
the following samples’ results were both wrongly predicted as “Real Samples”. On the
other hand, some real samples were wrongly classified as “Fake Samples", which is not
explainable by current detection algorithms. Thus, another challenge of deepfake detection
is required to increase the precision and false-positive rate as well. Accuracy is a common
metric but not the best evaluating metric to rate a model’s performance.

Figure 14. Some fake samples with obvious forgery clues are wrongly predicted as “Real”.

4. Conclusions and Challenges

This survey offers a timely summary of deepfake detection datasets with their ad-
vantages and disadvantages and classifies deepfake detection methods into three com-
monly used categories from the perspectives of feature extraction methods. Moreover, we
performed several in-dataset and cross-dataset testing experiments to calculate relative
evaluating metrics such as accuracy and AUC by reproducing Capsule Net, CORE, and
STIL codes. From the investigation, we propose three conclusions:
1. It is difficult for traditional CNN-based methods to provide good generalization ability
on unseen data. This is because different forgery methods provide data qualities. For
more realistic manipulation, it is difficult for the current model to distinguish from
the texture level.
2. Even though semi-supervised methods also utilize the CNN-based backbone to extract
features, they put more effort into calculating representation similarity and patch
similarity or analyzing the temporal inconsistency of videos. They focus on the data
themselves and find forgery clues from artifacts by different data augmentations or by
the inconsistency of spatial and temporal information, which is an effective detection
method in the current stage.
3. Frame-based detection also has another drawback in that it performs worse on Neural
Textures because of the improvement in visual effects and imitating realistic facial
structures. The Neural Textures algorithm can ensure the three-dimensional consis-
tency of generative images, which means that an image rendered at different viewing
Electronics 2024, 13, 585 19 of 22

angles is similar to the observation of real, three-dimensional human faces. This

forgery method is similar to the inpainting task to let AI fill in the defective parts to
synthesize high-quality fakes. Thus, it is necessary to explore temporal-based detec-
tion to find the inconsistent relationship between video clips to solve this limitation.
Vision transformer has already proven that it can perform better benchmark on image
classification tasks because it can provide the correlation between each patch feature, which
is hard to complete in the CNN backbone. We investigated the potential of transformer
blocks and assumed it could combine with consistency or inconsistency learning, for
example, the extracted features obtained from the transformer can directly feed to the
similarity calculation block by dividing the extracted features into several patches, and
then the loss function is more diverse to filter out more “hard” samples. This aim is to find
the inconsistency relationship between patches’ PRNU noises, which can be an effective
method to detect deepfakes. Since in the video transformer developed the attention is
not only on different spatial features anymore, it can focus on temporal inconsistency by
reshaping extracted features.
In addition, the main challenge of the transformer block is it requires corresponding
pre-trained models. We utilized the transformer encoder block with initialized parameters
and cross-entropy loss on the FF++ training set to train. It always showed low accuracy,
reaching 64% on the validating phase, which is even worse than Capsule Network. Thus, it
is important to select well-trained model parameters with the corresponding model archi-
tecture. Finally, we additionally propose three ideas for future deepfake detection trends:
1. For frame-based detection methods, consistency and inconsistency learning will
become mainstream by combining transformer blocks. The total loss is also under
the guidance of classification loss, consistency loss, and inconsistency loss, and the
backbone of feature extraction can be replaced by transformer architectures to obtain
a better corresponding feature.
2. Not only can deepfake detection detect forgery clues via videos, but detection mod-
els will develop multi-modalities such as the inconsistency of audio and detecting
manipulated source generators. This is because of the development of biological
signal-based detectors. They fuse biological signals such as PPG and provide and
generate multi-modal data from different previous computer vision methods, which
greatly enriches the diversity of forgery information and the model’s learning ability.
3. There is still a lot of room to develop spatial and temporal detection methods with dif-
ferent methods to extract and fuse features. Video transformers can play a significant
role in performing spatial and temporal feature extraction and attention mechanisms
in this field.
Altogether, researchers face two main challenges in future deepfake detection. Firstly,
the biggest challenge is the lack of datasets with large scale, various quality levels, and
multiple attack methods. It is significant to enrich the diversity of data, including different
ethnicities and shooting scenarios at the current stage, because models can learn different
forgery features from data level while training, and it will be the first step to solving the
problem of model generalization ability. Thus, datasets should develop with the deepfake
generation development and maintain a long-term attacker and defender competition.
Secondly, most forgery detectors are based on video frames. There is a lack of inter-frame
temporal consistency detection methods to solve the problems of temporal abnormalities
and real/fake frames in consecutive intervals. Finally, the ethical implications of deepfake
detection will be discussed. As both generation and detection technologies continue to
evolve in a competitive relationship, it is difficult to judge the ethics of deepfake detectors
in the current stage. However, it can be concluded that unsophisticated detection skills and
highly resourced sponsors’ misinformation sometimes are at an intersection, which often
misleads technical updates and ethics in this field. Thus, there is no “easy” approach to
navigate the detector dilemma, but a set of implications derived from multi-stakeholders
can better inform detection to process decisions and policy in the practice.
Electronics 2024, 13, 585 20 of 22

Author Contributions: Conceptualization, L.Y.G. and X.J.L.; methodology, L.Y.G. and X.J.L.; software,
L.Y.G.; validation, L.Y.G.; formal analysis, L.Y.G.; investigation, L.Y.G.; resources, L.Y.G. and X.J.L.;
data curation, L.Y.G. and X.J.L.; writing—original draft preparation, L.Y.G. and X.J.L.; writing—
review and editing, L.Y.G. and X.J.L.; visualization, L.Y.G. and X.J.L.; supervision, X.J.L.; project
administration, X.J.L. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Data Availability Statement: The relative datasets are downloaded from [8,13]. The reproduced
codes are from [3,11,15,26,27,53].
Conflicts of Interest: The authors declare no conflicts of interest.

References
1. Abdulreda, A.S.; Obaid, A.J. A landscape view of deepfake techniques and detection methods. Int. J. Nonlinear Anal. Appl. 2022,
13, 745–755.
2. Zhang, L.; Lu, T. Overview of Facial Deepfake Video Detection Methods. J. Front. Comput. Sci. Technol. 2022, 17, 1–26.
3. FaceSwap-GAN. Available online: https://github.com/shaoanlu/faceswap-GAN (accessed on 15 December 2018).
4. Goodfellow, I.; Pouget-Abadie, J.; Mirza, M.; Xu, B.; Wared-Farley, D.; Ozair, S.; Courville, A.; Bengio, Y. Generative Adversarial
Network. Proceeding Commun. ACM 2018, 63, 139–144. [CrossRef]
5. Radford, A.; Metz, L.; Chintala, S. Unsupervised representation learning with deep convolutional generative adversarial networks.
In Proceedings of the Computer Vision and Pattern Recognition, Boston, MA, USA, 7–12 June 2015.
6. Ho, J.; Jain, A.; Abbeel, P. Denoising Diffusion Probabilistic Models. In Proceedings of the 34th International Conference on
Neural Information Processing System, Red Hook, NY, USA, 6–12 December 2020; pp. 6840–6851.
7. Rana, M.S.; Nobi, M.N.; Murali, B.; Sung, A.H. Deepfake Detection: A Systematic Literature Review. IEEE Access 2022, 10,
25494–25513. [CrossRef]
8. Kaggle. Available online: https://www.kaggle.com/c/deepfake-detection-challenge/overview (accessed on 12 December 2023).
9. Dimensions Scholarly Database. Available online: https://app.dimensions.ai/ (accessed on 10 December 2023).
10. DeepFaceLive. Available online: https://github.com/iperov/DeepFaceLive (accessed on 9 November 2023).
11. Roop. Available online: https://github.com/s0md3v/roop (accessed on 11 October 2023).
12. Li, Y.Z.; Yang, X.; Sun, P.; Qi, H.G.; Lyu, S. Celeb-DF: A Large-scale Challenging Dataset for DeepFake Forensics. In Proceedings
of the Computer Vision and Pattern Recognition, Seattle, WA, USA, 14–19 June 2020.
13. Zhou, T.F.; Wang, W.G.; Liang, Z.Y.; Shen, J.B. Face Forensics in the Wild. In Proceedings of the Computer Vision and Pattern
Recognition (CVPR), Virtual, 19–25 June 2021.
14. Guo, J.; Deng, J.; Lattas, A.; Zafeirioul, S. Sample and Computation Redistribution for Efficient Face Detection. In Proceedings of
the Computer Vision and Pattern Recognition, Seattle, WA, USA, 14–19 June 2020.
15. FaceSwap. Available online: https://github.com/deepfakes/faceswap (accessed on 10 November 2020).
16. Tolosana, R.; Romero-Tapiador, S. DeepFakes Evolution: Analysis of Facial Regions and Fake Detection Performance. In
Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Seattle, WA, USA, 14–19 June 2020.
17. Dang, H.; Liu, F.; Stehouwer, J.; Liu, X.; Jain, A. On the Detection of Digital Face Manipulation. In Proceedings of the IEEE/CVF
Conference on Computer Vision and Pattern Recognition, Seattle, WA, USA, 14–19 June 2020.
18. Karras, T.; Laine, S.; Aila, A. A Style-Based Generator Architecture for Generative Adversarial Networks. In Proceedings of the
Computer Vision and Pattern Recognition, Long Beach, CA, USA, 16–20 June 2019.
19. Li, Y.; Chang, M.C.; Lyu, S. In Ictu Oculi: Exposing AI Created Fake Videos by Detecting Eye Blinking. In Proceedings of the IEEE
International Workshop on Information Forensics and Security, Hong Kong, China, 11–13 December 2018; pp. 1–7.
20. Nguyen, T.T.; Nguyen, Q.; Nguyen, D.; Nguyen, D.T.; Huynh-The, T.; Nahavandi, S.; Nguyen, T.; Pham, Q.; Nguyen, C. Deep
Learning for Deepfakes Creation and Detection: A Survey. arXiv 2022, arXiv:1909.11573.
21. Alahamari, F.; Naim, A.; Alqahtani, H. IoT-enabled Convolutional Neural Networks: Techniques and Applications. Chap-
ter: E-Learning Modelling Technique and Convolution Neural Networks in Online Education. 2023. Available online:
https://www.taylorfrancis.com/chapters/edit/10.1201/9781003393030-10/learning-modeling-technique-convolution-neural-
networks-online-education-fahad-alahmari-arshi-naim-hamed-alqa (accessed on 5 January 2024).
22. Sabour, S.; Frosst, N.; Hinton, G.E. Dynamic Routing between Capsules. In Proceedings of the 31st International Conference on
Neural Information Processing Systems, Long Beach, CA, USA, 4–9 December 2017; pp. 3859–3869.
23. Nguyen, H.H.; Yamagishi, J.; Echizen, I. Use of a Capsule Network to Detect Fake Images and Videos. In Proceedings of the
Computer Vision and Pattern Recognition, Long Beach, CA, USA, 16–20 June 2019.
24. Simonyan, K.; Zisserman, A. Very Deep Convolutional Networks for Large-Scale Image Recognition. In Proceedings of the
Computer Vision and Pattern Recognition, Columbus, OH, USA, 23–28 June 2014.
Electronics 2024, 13, 585 21 of 22

25. ImageNet Large Scale Visual Recognition Challenge (ILSVRC). Available online: https://image-net.org/challenges/LSVRC/
(accessed on 13 December 2023).
26. Capsule-Forensics-v2: Implementation of the Capsule-Forensics-v2. Available online: https://github.com/nii-yamagishilab/
Capsule-Forensics-v2 (accessed on 29 October 2019).
27. Ni, Y.; Meng, D.; Yu, C.; Quan, C.B.; Ren, D.; Zhao, Y. CORE: Consistent Representation Learning for Face Forgery Detection. In
Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, LA, USA, 18–24 June 2022.
28. Zhong, Z.; Zheng, L.; Kang, G.; Li, S.; Yang, Y. Random Erasing data argumentation. In Proceedings of the AAAI, New York, NY,
USA, 7–12 February 2020.
29. Chollet, F. Xception: Deep learning with depthwise separable convolutions. In Proceedings of the Computer Vision and Pattern
Recognition, Honolulu, HI, USA, 21–26 July 2017.
30. Li, L.; Bao, J.; Zhang, T.; Yang, H.; Chen, D.; Wen, F.; Guo, B. Face X-ray for More General Face Forgery Detection. In
Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA, 13–19 June
2020; pp. 5000–5009.
31. Li, Y.; Lyu, S. Exposing DeepFake Videos By Detecting Face Warping Artifacts. In Proceedings of the IEEE/CVF Conference on
Computer Vision and Pattern Recognition Workshops, Long Beach, CA, USA, 16–17 June 2019; pp. 16–20.
32. Zhao, T.; Xu, X.; Xu, M.; Ding, H.; Xiong, Y.; Xia, W. Learning Self-Consistency for Deepfake Detection. In Proceedings of the
IEEE/CVF Conference on Computer Vision and Pattern Recognition, Nashville, TN, USA, 20–25 June 2021.
33. Chen, M.; Fridrich, J.; Goljan, M.; Lukas, J. Determining Image Origin and Integrity Using Sensor Noise. IEEE Trans. Inf. Forensics
Secur. 2008, 3, 74–90. [CrossRef]
34. Barni, M.; Bondi, L.; Bonettini, N.; Bestagini, P.; Costanzo, A.; Maggini, M.; Tondi, B.; Tubaro, S. Aligned and Non-Aligned Double
JPEG Detection Using Convolutional Neural Networks. J. Vis. Commun. Image Represent. 2017, 49, 153–163. [CrossRef]
35. He, K.; Zhang, X.; Ren, S.; Sun, J. Deep Residual Learning for Image Recognition. In Proceedings of the 2016 IEEE Conference on
Computer Vision and Pattern Recognition, Las Vegas, NV, USA, 27–30 June 2016.
36. Sun, K.; Yao, T.; Chen, S.; Ding, S.; Li, J.; Ji, R. Dual Contrastive Learning for General Face Forgery Detection. In Proceedings of
the AAAI Conference on Artificial Intelligence, Washington, DC, USA, 7–14 February 2022; Volume 36, pp. 2316–2324.
37. Fridrich, J.; Kodovsky, J. Rich Models for Steganalysis of Digital Images. IEEE Trans. Inf. Forensics Secur. 2012, 7, 868–882.
[CrossRef]
38. Gutmann, M.; Hyvarinen, A. Noise-contrastive estimation: A new estimation principle for unnormalized statistical models. In
Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, Sardinia, Italy, 13–15 May 2010;
Volume 9, pp. 297–304.
39. Shi, X.; Chen, Z.; Wang, H.; Yeung, D. Convolutional LSTM Network: A Machine Learning Approach for Precipitation Network.
In Proceedings of the Advances in Neural Information Processing Systems, Montreal, QC, Canada, 7–12 December 2015.
40. Gu, Z.; Chen, Y.; Yao, T.; Ding, S.; Li, J.; Huang, F.; Ma, L. Spatiotemporal Inconsistency Learning for Deepfake Video Detection.
In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Virtual, 19–25 June 2021.
41. Dosovitskiy, A.; Beyer, L.; Kolesnikov, A.; Weissenborn, D.; Zhai, X.; Unterthiner, T.; Dehghani, M.; Minderer, M.; Heigold, G.;
Gelly, S.; et al. An Image is Worth 16 × 16 Words: Transformers for Image Recognition at Scale. In Proceedings of the International
Conference on Learning Representations, Virtual, 26 April–1 May 2020.
42. Khan, S.A.; Dai, H. Video Transformer for Deepfake Detection with Incremental Learning. In Proceedings of the 29th ACM
International Conference on Multimedia, Virtual Event, 20–24 October 2021; pp. 1821–1828.
43. Wodajo, D.; Atnafu, S. Deepfake Video Detection Using Convolutional Vision Transformer. In Proceedings of the Computer
Vision and Pattern Recognition, Nashville, TN, USA, 19–25 June 2021.
44. Khormali, A.; Yuan, J. DFDT: An End-to-End DeepFake Detection Framework Using Vision Transformer. Appl. Sci. 2022, 12, 2953.
[CrossRef]
45. Trockman, A.; Zico Kolter, J. Patches Are All You Need? In Proceedings of the Computer Vision and Pattern Recognition, New
Orleans, LA, USA, 18–24 June 2022.
46. Zhao, C.; Wang, C.; Hu, G.; Chen, H.; Liu, C.; Tang, J. ISTVT: Interpretable Spatial-Temporal Video Transformer for Deepfake
Detection. IEEE Trans. Inf. Forensics Secur. 2023, 18, 1335–1348. [CrossRef]
47. Shaw, P.; Uszkoreit, J.; Vaswani, A. Self-attention with Relative Position Representations. In Proceedings of the NAACL 2018,
New Orleans, LA, USA, 1–6 June 2018.
48. Ciftci, U.A.; Demir, İ.; Yin, L. How Do the Hearts of Deep Fakes Beat? Deep Fake Source Detection via Interpreting Residuals
with Biological Signals. In Proceedings of the IEEE International Joint Conference on Biometrics (IJCB), Houston, TX, USA, 8
September–1 October 2020; pp. 1–10.
49. Wu, J.; Zhu, Y.; Jiang, X.; Liu, Y.; Lin, J. Local attention and long-distance interaction of rPPG for deepfake detection. In
Proceedings of the Visual Computer, Lake Tahoe, NV, USA, 16–18 October 2023.
50. Stefanov, K.; Paliwal, B.; Dhall, A. Visual Representation of Physiological Signals for Fake Video Detection. In Proceedings of the
IEEE/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, LA, USA, 18–24 June 2022.
51. Kipf, T.N.; Welling, M. Semi-Supervised Classification with Graph Convolutional Networks. In Proceedings of the 5th Interna-
tional Conference on Learning Representations (ICLR 2017), Toulon, France, 24–26 April 2017.
Electronics 2024, 13, 585 22 of 22

52. Liu, X.; Fromm, J.; Patel, S.N.; McDuff, D. Multi-task temporal shift attention networks for on-device contactless vital measure-
ments. Adv. Neural Inf. Process. Syst. 2020, 33, 19400–19411.
53. DFDC-Selium. Available online: https://github.com/selimsef/dfdc_deepfake_challenge (accessed on 11 December 2022).

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.