A Novel ResNeXt-based CNN for Spatial Feature Extraction and  LSTM based RNN for Temporal Pattern Analysis of Deepfake Detection Systems

Ashima Gajendra Singh,Dr. Pooja Sharma

A Novel ResNeXt-based CNN for Spatial Feature Extraction and LSTM based RNN for Temporal Pattern Analysis of Deepfake Detection Systems

Authors

Ashima Gajendra Singh,Dr. Pooja Sharma

Keywords:

Deepfake Detection, Generative Adversarial Networks (GANs), Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), Hybrid CNN-LSTM Model, Adversarial Robustness, Digital Media Forensics.

Abstract

Deepfake media, generated using advanced AI-based techniques, pose a significant threat toinformation integrity and public trust. This paper presents a novel deep learning frameworkfor deepfake detection, combining the strengths of Convolutional Neural Networks (CNNs)and Recurrent Neural Networks (RNNs). The proposed method employs a ResNeXt-based CNN architecture to effectively

References

Generative Adversarial Networks (GANs) – Ian Goodfellow et al., "Generative Adversarial Networks," Advances in Neural Information Processing Systems (NeurIPS), 2014.

StyleGAN & Face Manipulation – Karras, T., Laine, S., Aila, T., "A Style-Based Generator Architecture for Generative Adversarial Networks," CVPR, 2019.

Downloads

Published

2024-02-20

How to Cite

Ashima Gajendra Singh,Dr. Pooja Sharma. (2024). A Novel ResNeXt-based CNN for Spatial Feature Extraction and LSTM based RNN for Temporal Pattern Analysis of Deepfake Detection Systems. Journal of Computational Analysis and Applications (JoCAAA), 33(2), 1811–1826. Retrieved from https://www.eudoxuspress.com/index.php/pub/article/view/3570