Face swapping based on machine learning

• Main Author: Zhou, Suxi
• Other Authors: Jiang Xudong
• Format: Thesis-Master by Coursework
• Language: English
• Published: Nanyang Technological University 2021
• Subjects: Engineering::Computer science and engineering::Information systems::Information systems applications; Engineering::Electrical and electronic engineering
• Online Access: https://hdl.handle.net/10356/150319
• Institution: Nanyang Technological University

Out of the increasing demand of internet security and entertainment, the face swapping technology attracts great attention in both academic area and commercial companies. This dissertation mainly construct a face swapping system. Firstly use Histogram of Oriented Gradient (HOG) to detect the face in a given image, and use training data to generate a prediction model based on the gradient boosting decision tree (GBDT) algorithm to extract the coordinates of 81 feature points of facial features and facial contours; Then train the Multi-Layer Perceptron (MLP) classifier to predict the gender and race of the face to be recognized and find the reference face image of the same gender race. Lastly, use the extracted feature point coordinates to exchange the facial features of the target face image and the reference face image.