Welcome to the upgraded BRAC University Institutional Repository. We are currently organizing collections after a recent system upgrade. Homepage category counters may temporarily show lower numbers while syncing, but over 27,000 repository items remain safe and accessible. Please use the search bar to find theses, scholarly outputs, and institutional documents.

Deepfake detection using neural networks

Citation

Abstract

Deepfake is a sort of arti cial intelligence that forge original image or video and create persuading images, audio and video hoaxes by utilizing two contending AI algorithms-the generator and discriminator that form a generative adversarial net- work (GAN). The term `Deepfake' started in 2017, when a mysterious Reddit user called himself "Deepfakes." The user "Deepfakes" supplanted genuine faces with celebrity faces. With the rapid advancement of modern technology, Deepfakes have become an emerging problem, as deepfakes can threaten cybersecurity, political elec- tions, companies, individual and corporate nances, reputations, and more. There- fore, this makes deepfake detection more and more urgent. Although, a lot of techniques has been invented to detect deepfake but not all of them works perfectly and accurately for all cases. Also, as more up to date deepfake creation strategies are grown, ine ectively generalizing methodologies should be continually refreshed to cover these new techniques. Our research focuses on the recent techniques that are used to create manipulated videos and detect them though ensembling di erent CNN models.

Description

Cataloged from PDF version of thesis.
Includes bibliographical references (pages 33-34).
This thesis is submitted in partial fulfillment of the requirements for the degree of Bachelor of Science in Computer Science and Engineering, 2021.

Publisher Link

Type

Thesis