Enhanced AI: Deepfake Detection System
DOI:
https://doi.org/10.63282/3050-922X.ICAILLMBA-104Keywords:
Deepfake Detection, ResNext-50, LSTM, Temporal Analysis, AI, Real-Time Video ProcessingAbstract
With the rapid growth of generative artificial intelligence, deepfakes have become a serious concern due to their potential to mislead, manipulate, and harm individuals and communities. This paper presents a real-time AI solution that integrates ResNext-50 convolutional networks with Long Short-Term Memory (LSTM) networks to detect synthetic videos with high accuracy. The model identifies spatial artifacts and inconsistencies in temporal dynamics across frames. Trained on a balanced dataset of 6,000 videos (both authentic and fake), the system achieves 97.76% accuracy and is optimized for real-time usage with a practical deployment interface. This work is especially relevant for social media moderation, forensic analysis, and digital content verification.
References
[1] Rossler et al. (2019), "FaceForensics++", arXiv:1901.08971 Li et al. (2020), "Celeb-DF", arXiv:1909.12962
[2] Güera & Delp (2018), IEEE AVSS
[3] Chugh et al. (2020), Audio-Visual Deepfake Detection Khalid et al. (2021), CNN-LSTM Hybrid Detection
[4] Yuezun Li, Siwei Lyu, “ExposingDF Videos By Detecting Face Warping Arti- facts,” in arXiv:1811.00656v3. Yuezun Li, Ming-Ching Chang and Siwei Lyu “Exposing AI Created Fake Videos by Detecting Eye Blinking” in arXiv:1806.02877v2.
[5] Huy H. Nguyen , Junichi Yamagishi, and Isao Echizen “ Using capsule net- works to detect forged images and videos ” in arXiv:1810.11215.
[6] D. Güera and E. J. Delp, "Deepfake Video Detection Using Recurrent Neural Networks," 2018 15th IEEE International Conference on Advanced Video and Sig- nal Based Surveillance (AVSS), Auckland, New Zealand, 2018, pp. 1-6.
[7] Laptev, M. Marszalek, C. Schmid, and B. Rozenfeld. Learning realistic hu- man actions from movies. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 1–8, June 2008. Anchorage, AK
[8] Umur Aybars Ciftci, İlke Demir, Lijun Yin “Detection of Synthetic Portrait Videos using Biological Signals” in arXiv:1901.02212v2
[9] D. P. Kingma and J. Ba. Adam: A method for stochastic optimization. arXiv:1412.6980, Dec. 2014. ResNext Model : https://pytorch.org/hub/pytorch_vision_resnext/ accessed on 06 April 2020 https://www.geeksforgeeks.org/software-engineering-cocomo-model/ Accessed on 15 April 2020 Deepfake Video Detection using Neural Networks http://www.ijsrd.com/articles/IJSRDV8I10860.pdf International Journal for Scientific Research and Development http://ijsrd.com/
