Affiliation:
1. Nanyang Technological University Singapore
Abstract
ABSTRACTDeepfake research has gained traction in recent years. Surveys have been conducted to summarize work on the detection and generation of deepfakes. However, a more comprehensive and quantitative overview that encompasses both technical and non‐technical areas is lacking. We address this gap using topic modelling to discover deepfake research topics in academic publications. Our results show that while detection techniques topics dominate the research field, other areas, such as privacy and legal research, offer potential avenues for further exploration.
Subject
Library and Information Sciences,General Computer Science
Reference8 articles.
1. Latent dirichlet allocation;Blei D. M.;Journal of Machine Learning Research,2003
2. Deepfakes on Twitter: Which Actors Control Their Spread?
3. A topic modeling comparison between lda, nmf, top2vec, and bertopic to demystify twitter posts. Frontiers;Egger R.;Sociology,2022
4. Deepfakes: Trick or treat?
5. Determining authenticity of video evidence in the age of artificial intelligence and in the wake of Deepfake videos