Year: 2024
Type: article
Abstract: It has become more difficult to detect the difference between real and false media because to the rapid advancements in computer graphics and highly artificial intelligence in the production of realistic photos and videos in recent years. complicated. Even if these computer-generated images or movies have practical uses, they can also present privacy and security risks. Deepfake is one method that carries these hazards. Combining the terms "deep learning" and "fake" yields the term "deepfake." Anyone can use Deepfake to edit or erase another person's face from a photo or video. The speech and facial emotions of an original image or video can likewise be altered using deepfakes. These days, deepfake algorithms use artificial intelligence and deep learning to replace the original voice, face, or emotions. It's difficult to tell whether the content has been modified by deepfake techniques. Deepfakes are altered and retouched using deep learning algorithms, which makes it more difficult to distinguish between actual and fake images. Generative adversarial neural networks (GANs) are used to create deepfakes, which may be dangerous for the public. It's imperative to spot fake graphic content with great attention. Several investigations have been carried out to identify deepfakes in photo manipulation. The two main problems with the existing approaches are their time consumption and accuracy.
Author: Praful Bharadiya, Jasmine
Institution: University of the Cumberlands
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W7218330720
