EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

CSTAN: A Deepfake Detection Network with CST Attention for Superior Generalization

Rui Yang iD, Kang You, Cheng Pang, Xiaonan Luo iD, Rushi Lan iD

Sensors · 2024

DOI: 10.3390/s24227101 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, 2024. Verfasst von Rui Yang und 4 weiteren Autor:innen. Beteiligte Einrichtung: Guangxi Key Laboratory of Image and Graphic Intelligent Processing, Guilin University of Electronic Technology, Guilin 541004, China. Hauptthemen laut MEDLINE: Neural Networks, Computer, Deep Learning. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 24 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: the Guangxi Science and Technology Project, the National Natural Science Foundation of China, the Guangxi Natural Science Foundation.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

With the advancement of deepfake forgery technology, highly realistic fake faces have posed serious security risks to sensor-based facial recognition systems. Recent deepfake detection models mainly use binary classification models based on deep learning. Despite achieving high detection accuracy on intra-datasets, these models lack generalization ability when applied to cross-datasets. We propose a deepfake detection model named Channel-Spatial-Triplet Attention Network (CSTAN), which focuses on the difference between real and fake features, thereby enhancing the generality of the detection model. To enhance the feature-learning ability of the model for image forgery regions, we have designed the Channel-Spatial-Triplet (CST) attention mechanism, which extracts subtle local information by capturing feature channels and the spatial correlation of three different scales. Additionally, we propose a novel feature extraction method, OD-ResNet-34, by embedding ODConv into the feature extraction network to enhance its dynamic adaptability to data features. Trained on the FF++ dataset and tested on the Celeb-DF-v1 and Celeb-DF-v2 datasets, the experimental results show that our model has stronger generalization ability in cross-datasets than similar models.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Neural Networks, ComputerDeep LearningHumansAlgorithmsAutomated Facial RecognitionImage Processing, Computer-AssistedFaceFacial Recognition

Stichwörter der Autor:innen

attention mechanismdeepfake detectiondetection modelfeature extraction

Autor:innen und Institutionen

  • Guangxi Key Laboratory of Image and Graphic Intelligent Processing, Guilin University of Electronic Technology, Guilin 541004, China.
  • School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin 541004, China.
  • International Joint Research Laboratory of Spatio-Temporal Information and Intelligent Location Services, Guilin University of Electronic Technology, Guilin 541004, China.

ORCID-Profile: Rui Yang, Xiaonan Luo, Rushi Lan

Förderung

  • the Guangxi Science and Technology Project
  • the National Natural Science Foundation of China
  • the Guangxi Natural Science Foundation
  • the Guangxi Key Laboratory of Image and Graphic Intelligent Processing Project
  • the Innovation Project of GUET Gurduate Education
  • Guangxi Science and Technology Project
  • National Natural Science Foundation of China
  • Guangxi Natural Science Foundation

Zitieren

APA

Yang, R., You, K., Pang, C., Luo, X., & Lan, R. (2024). CSTAN: A Deepfake Detection Network with CST Attention for Superior Generalization. Sensors. https://doi.org/10.3390/s24227101

Vancouver

Yang R, You K, Pang C, Luo X, Lan R. CSTAN: A Deepfake Detection Network with CST Attention for Superior Generalization. Sensors. 2024. doi:10.3390/s24227101

BibTeX
@article{Yang2024,
  author = {Yang, Rui and You, Kang and Pang, Cheng and Luo, Xiaonan and Lan, Rushi},
  title = {CSTAN: A Deepfake Detection Network with CST Attention for Superior Generalization},
  journal = {Sensors},
  year = {2024},
  doi = {10.3390/s24227101}
}
Kostenlos starten →