DeepFake-Detection-System

DeepFake Detection System using CNN and OpenCV This Deep learing model uses CNN algorithm to perform video analytics and detect whether the test video is a computer generated deepfake or not.

We use various packages and libraries like Tensorflow, Keras, OpenCV and imageIO to train the neural network model and test on a set of data which consists of both, deepfake and also videos that are not synthetically generated

Then the model is tested on a testing dataset through which we can classify if the video file is a deepfake or not.

Requirements:

1. Python 3.11
2. Tensorflow 2.0
3. Keras
4. OpenCV2
5. Matplotlib Video dataset from Kaggle- https://www.kaggle.com/code/krooz0/deep-fake-detection-on-images-and-videos/input?select= https://www.kaggle.com/code/krooz0/deep-fake-detection-on-images-and-videos/input?select=train_sample_videos

Steps to run the program:

1. Install the required softwares and packages in your prefered python compiler.
2. Download the dataset from Kaggle Deepfake Detection Challenge.
3. Create a python workspace and import your dataset into the workspace.
4. Make sure to read the data frame appropriately.
5. Run the code
6. Test it on new testing data from else where to evaluate the model performance in a real time scenario.