Abstract
INTRODUCTION: The rapid progress in artificial intelligence (AI) over the past decade has ushered in a new era of transformative technologies. Deep learning has emerged as a potential tool, demonstrating remarkable capabilities in various applications. This paper focuses on one of the controversial applications of deep learning commonly known as deepfakes.OBJECTIVES: The main objective of this comprehensive bibliometric survey is to explore the trends, applications and challenges of deepfakes over the course of last 4.5 years.METHODS: In this research, a total of 794 documents published from 2019 to July 2023 were acquired from Scopus database. To conduct this bibliometric analysis, RStudio and VOSviewer tools have been used. In this current analysis, deepfake challenges, countries, sources, top 20 cited documents, and research trends in the field of deepfake have been included.RESULTS: The analysis highlights a substantial increase in deepfake publications from January 2019 to July 2023. Out of the 8 document types identified 38% are article publications. In addition, from the journal articles it has been depicted that the journal source entitled "Advances in Computer Vision and Pattern Recognition" holds Q1 status with 8.3% publications in the deepfakes domain during the targeted year range. Moreover, the data visualizations reveal the growing international collaboration, with the USA as the most prolific country in deepfake research.CONCLUSION: Despite numerous reviews on deepfakes, there has been a notable absence of comprehensive scientometric analyses. This paper fills this gap through a bibliometric study using the Scopus database as urderlying source. The analysis includes keyword analysis, leading research-contributing institutes, co-country collaboration, and co-keyword occurrence. The findings offer valuable insights for scholars, providing a foundational understanding including document types, prominent journals, international collaboration trends, and influential institutions and offering valuable guidance for future scholarly pursuits in this evolving field.
Publisher
European Alliance for Innovation n.o.
Reference85 articles.
1. Leandro A Passos, Danilo Jodas, Kelton AP da Costa, Luis A Souza Júnior, Douglas Rodrigues, Javier Del Ser, David Camacho, and João Paulo Papa. A review of deep learning-based approaches for deepfake content detection. arXiv preprint arXiv:2202.06095, 2022.
2. Abdulqader M Almars. Deepfakes detection techniques using deep learning: a survey. Journal of Computer and Communications, 9(05):20–35, 2021.
3. P Marcel. Deepfakes: a new threat to face recognition. Assessment and detection, 2018.
4. Ai-powered deepfakes bare fangs in 2023, raise concern about impact on privacy, electoral politics — thehindubusinessline.com. https://www.thehindubusinessline.com/info-tech/ai-powered-deepfakes-bare-fangs-in-2023-raise-concern-about-impact-on-privacy-electoral-politics/article67692281.ece, 2023. [online; accessed 01-january-2024].
5. Neha Sandotra and Bhavna Arora. A comprehensive evaluation of feature-based ai techniques for deepfake detection. Neural Computing and Applications, pages 1– 29, 2023.