A Survey on Deepfake Analysis and Recognition using Deep Learning
| Main Authors: | Deepak N R, Mohamed Aweez Akram, Mohammed Adnan K, Mohammed Shibil, Shameema Banu R |
|---|---|
| Format: | Article Journal |
| Bahasa: | eng |
| Terbitan: |
, 2021
|
| Subjects: | |
| Online Access: |
https://zenodo.org/record/5040481 |
Daftar Isi:
- The recent advancements in the field of artificial intelligence, deep learning and image processing, has led to the rise of a software called deepfake. It is a tool that can produce extreme transformations in human faces like aging, gender swap, etc., and can make someone say and do things which never happened in reality. The resulting content produced are utterly realistic, which can be highly dangerous and may have the potential of altering the truth and eroding trust by giving false reality. Deepfakes can have negative or positive implications on society. It can be used in different domains like advertising, creative arts, film production, video games, etc., to name a few. But it could also pose huge security threats like influence the public opinion during elections, perpetrate fraud or blackmail someone. Current solutions can be used to recognize only specific manipulation techniques like splicing, colouring, etc., and results provided have poor accuracy. Hence, there is a need for automated tools competent of detecting fake multimedia content.