DeepFake Detection

Implementation for DeepFake Detection Using Ensembling Techniques

Determining whether a given video is Real or Fake by cropping the face of a person and classifying the cropped image by ensembling ResNext and EfficientNetB6

Installation

1. Clone this repository.

https://github.com/kenil-shah/DeepFake_Detection.git

2. In the repository, execute pip install -r requirements.txt to install all the necessary libraries.

3. Three deep learning models are used inorder to determine the class of video

   1. YOLO Face model:- Used to determine the coordinates of the face of person and generate a cropped facial image using those coordinates
   2. ResNext:- First Model used for ensembling
   3. EfficientNetB6(With Attention):- Second Model used for ensembling

4. Download the pretrained weights.

   1. YOLO Face model pretrained weights and save it in /model_data/
   2. ResNext:- pretrained weights and save it in the root directory
   3. EfficientNetB6(With Attention)pretrained weight and save it in the root directory