Deepfake Detection: Current Challenges and Next Steps

@article{Lyu2020DeepfakeDC,
  title={Deepfake Detection: Current Challenges and Next Steps},
  author={Siwei Lyu},
  journal={2020 IEEE International Conference on Multimedia \& Expo Workshops (ICMEW)},
  year={2020},
  pages={1-6}
}
  • Siwei Lyu
  • Published 11 March 2020
  • Computer Science, Physics
  • 2020 IEEE International Conference on Multimedia & Expo Workshops (ICMEW)
High quality fake videos and audios generated by AI- algorithms (the deep fakes) have started to challenge the status of videos and audios as definitive evidence of events. In this paper, we highlight a few of these challenges and discuss the research opportunities in this direction. 

Figures from this paper

The Presidential Deepfakes Dataset

TLDR
This work presents the Presidential Deepfakes Dataset (PDD), which consists of 32 videos, half of which are original videos and half ofWhich are manipulated with audio impersonations, synthesized lip synchronizations, political misinformation, and situational artifacts, which expands the context on which end-to-end media forensic systems can be evaluated.

DEEPFAKE DETECTION TECHNIQUES: A REVIEW

TLDR
Various techniques that are used by several researchers to detect Deepfake videos, based on Machine Learning, Support Vector Machine and Deep Learning based techniques such as Convolution Neural Network with or without LSTM are analyzed.

DeepFakes: Detecting Forged and Synthetic Media Content Using Machine Learning

TLDR
This study presents challenges, research trends, and directions related to DeepFake creation and detection techniques by reviewing the notable research in the DeepFake domain to facilitate the development of more robust approaches that could deal with the more advance DeepFake in future.

Multimodal Approach for DeepFake Detection

TLDR
A multimodal deepfake detection solution submitted against the Facebook DeepFake Detection Challenge, a state of the art benchmark dataset and competition released at the end of 2019, and was ultimately ranked among the top 25% of teams.

Model Attribution of Face-swap Deepfake Videos

TLDR
This paper introduces a new dataset with DeepFakes from Different Models (DFDM) based on several Autoencoder models, takes Deepfakes model attribution as a multiclass classification task and proposes a spatial and temporal attention based method to explore the differences among Deepfake videos in the new dataset.

Deepfake Video Detection Based on EfficientNet-V2 Network

TLDR
This paper proposes using a new network of EfficientNet-V2 to distinguish the authenticity of pictures and videos, and highlights the superior performance of the new network by comparing the existing detection network with the actual training and testing results.

A Note on Deepfake Detection with Low-Resources

TLDR
This paper enhances MesoNet by replacing the original activation functions allowing a nearly 1% improvement as well as increasing the consistency of the results, and introduces and verified a new activation function - Pish that at the cost of slight time overhead allows even higher consistency.

DeepFake Detection using 3D-Xception Net with Discrete Fourier Transformation

TLDR
This work has proposed a new method for detecting deep fake videos using 3D Inflated Xception Net with Discrete Fourier Transformation, which is the first attempt to use a 3D Xception net for categorizing video-based data.

Deepfake detection through PRNU and logistic regression analyses

TLDR
This paper considers multiple approaches to distinguish between authentic and Deepfake videos, and discovers that none of the methods show definite proof of Deepfake presence or absence, and proposes a conceptual system to authenticate original media instead of discovering Deepfake forgeries.

A Qualitative Survey on Deep Learning Based Deep fake Video Creation and Detection Method

  • Computer Science
    Australian Journal of Engineering and Innovative Technology
  • 2022
TLDR
This paper gives a complete overview of deepfake approaches and supports the implementation of novel and more reliable methods to cope with the highly complicated deep fakes by studying the background of deepfakes and state-of-the-art deepfake detection methods.

References

SHOWING 1-10 OF 33 REFERENCES

Celeb-DF: A Large-Scale Challenging Dataset for DeepFake Forensics

TLDR
This work presents a new large-scale challenging DeepFake video dataset, Celeb-DF, which contains 5,639 high-quality DeepFake videos of celebrities generated using improved synthesis process and conducts a comprehensive evaluation of DeepFake detection methods and datasets to demonstrate the escalated level of challenges posed by Celebrity-DF.

The Deepfake Detection Challenge (DFDC) Preview Dataset

TLDR
A set of specific metrics to evaluate the performance have been defined and two existing models for detecting deepfakes have been tested to provide a reference performance baseline.

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

TLDR
This work describes a new method to expose fake face videos generated with neural networks based on detection of eye blinking in the videos, which is a physiological signal that is not well presented in the synthesized fake videos.

Recurrent-Convolution Approach to DeepFake Detection - State-Of-Art Results on FaceForensics++

TLDR
This work attempts to detect Deepfake, Face2Face and FaceSwap manipulations in videos with a recurrent approach and improves upon the previous state-of-the-art up to 4.55%.

MesoNet: a Compact Facial Video Forgery Detection Network

TLDR
A method to automatically and efficiently detect face tampering in videos, and particularly focuses on two recent techniques used to generate hyper-realistic forged videos: Deepfake and Face2Face.

Exposing DeepFake Videos By Detecting Face Warping Artifacts

TLDR
A new deep learning based method that can effectively distinguish AI-generated fake videos from real videos is described, which saves a plenty of time and resources in training data collection and is more robust compared to others.

DeepFakes: a New Threat to Face Recognition? Assessment and Detection

TLDR
This paper presents the first publicly available set of Deepfake videos generated from videos of VidTIMIT database, and demonstrates that GAN-generated Deep fake videos are challenging for both face recognition systems and existing detection methods.

Recurrent Convolutional Strategies for Face Manipulation Detection in Videos

TLDR
This work attempts to detect Deepfake, Face2Face and FaceSwap tampered faces in video streams using recurrent convolutional models and distill the best strategy for combining variations in these models along with domain specific face preprocessing techniques through extensive experimentation to obtain state-of-the-art performance on publicly available video-based facial manipulation benchmarks.

Capsule-forensics: Using Capsule Networks to Detect Forged Images and Videos

  • H. NguyenJ. YamagishiI. Echizen
  • Computer Science
    ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
  • 2019
TLDR
The method introduced in this paper uses a capsule network to detect various kinds of spoofs, from replay attacks using printed images or recorded videos to computer-generated videos using deep convolutional neural networks.

Deepfake Video Detection Using Recurrent Neural Networks

  • David GueraE. Delp
  • Computer Science
    2018 15th IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS)
  • 2018
TLDR
A temporal-aware pipeline to automatically detect deepfake videos is proposed that uses a convolutional neural network to extract frame-level features and a recurrent neural network that learns to classify if a video has been subject to manipulation or not.