Yefeng Yuan

Date of Award

11-3-2021

Document Type

Thesis - SCU Access Only

Publisher

Santa Clara : Santa Clara University, 2021.

Degree Name

Master of Science (MS)

Department

Computer Science and Engineering

First Advisor

Yuhong Liu

Abstract

Thanks to the rapid development of artificial intelligence algorithms like deep neural networks (DNNs), high quality fake videos and audios (Deepfakes) can be easily generated using pre-trained models and have been widely adopted in many popular applications. These Deepfakes have brought great challenges to the society by spreading fake information as evidence of events to mislead the public. However, with proper use, they can also have clear benefits in certain cases, such as movie production and art creation. In this thesis paper, state-of-the-art models to create and detect Deepfake videos are introduced and compared. Based on the latest work, a new model is proposed, which can improve existing methods to create more convincing Deepfake videos.

Download

Share

COinS