/Deepfake_detection_using_deep_learning

This projects aims in detection of video deepfakes using deep learning techniques like ResNext and LSTM. We have achived deepfake detection by using transfer learning where the pretrained ResNext CNN is used to obtain a feature vector, further the LSTM layer is trained using the features.

Primary LanguageHTMLGNU General Public License v3.0GPL-3.0

Deepfake detection using Deep Learning (ResNext and LSTM)

1. Introduction

This projects aims in detection of video deepfakes using deep learning techniques like ResNext and LSTM. We have achived deepfake detection by using transfer learning where the pretrained ResNext CNN is used to obtain a feature vector, further the LSTM layer is trained using the features.

2. System Architecture

3. Our Results

Model Name No of videos No of Frames Accuracy
model_84_acc_10_frames_final_data.pt 01 10 84.21461
model_87_acc_20_frames_final_data.pt 01 20 87.79160
model_89_acc_40_frames_final_data.pt 01 40 89.34681
model_90_acc_60_frames_final_data.pt 01 60 90.59097
model_91_acc_80_frames_final_data.pt 01 80 91.49818
model_93_acc_100_frames_final_data.pt 01 100 93.58794

4. Glimpse of our Application

  • Home page

  • Uploading video

  • Frames Splitting and cropping face video

  • Result

5. Supporting Documents

  • Presentation
  • Report
  • Abstract

6. Contributors

  • Mansi Mishra
  • Sanskriti Harmukh
  • Satyam Jain
  • Archit Chawda