Computer Vision: Python Face Swap and Quick Deepfake in Colab

Welcome to my new course Python face swap & quick deepface using google collab. Deepface is a world where what we see may not always be true. An example video will help us to define deep fake. Here it is. These videos were created using only one stamp-size photo of each person. You heard it right. You can make deepface of anyone with just a single image of your face and a few minutes using a normal computer. You can even make dead people speak or sing for you. Yes! Prepare to be amazed. Let me first explain the contents of this course. This course is divided into two halves. The first half will focus on creating a simple python-based face swap application. Before we move on, we will be introducing the deep fake technique and its advantages and disadvantages. Next, we need to prepare our computer with all dependencies installed. We will install anaconda, which is the platform and ide that we use for our Python program. There are a few sessions available later for those who wish to learn the basics about Python programming language basics. We will then install the remaining dependencies required to build our custom Python facial swap application. Than, we will write the Python program line by line. This will complete the program with over 300 lines. You can also download the entire code from the link to google drive provided in the last session. We will first be using two static images to do the face swap. One will be used as the source image, and the other as the target. Later, we will test it with real time video from our webcam. Than we will modify it so that it works with a pre-saved, saved video.