Redefining identity and security: Spoofing in the age of artificial intelligence
This research explores the growing threat of AI-driven deepfakes in the context of spoofing attacks. By systematically analyzing attack models, including facial reenactment, replacement, editing, and synthesis, as well as audio deepfake techniques, the study dissects both the creation mechanisms and...
- Autores:
-
Murillo Fonseca, Nicole
- Tipo de recurso:
- Trabajo de grado de pregrado
- Fecha de publicación:
- 2025
- Institución:
- Universidad de los Andes
- Repositorio:
- Séneca: repositorio Uniandes
- Idioma:
- eng
- OAI Identifier:
- oai:repositorio.uniandes.edu.co:1992/76116
- Acceso en línea:
- https://hdl.handle.net/1992/76116
- Palabra clave:
- Spoofing
Deepfake
Security
Detection
Authentication
Artificial intelligence
Ingeniería
- Rights
- openAccess
- License
- Attribution-NonCommercial 4.0 International
| Summary: | This research explores the growing threat of AI-driven deepfakes in the context of spoofing attacks. By systematically analyzing attack models, including facial reenactment, replacement, editing, and synthesis, as well as audio deepfake techniques, the study dissects both the creation mechanisms and detection methodologies underlying these sophisticated forms of edited media content. Leveraging insights from neural network architectures such as GANs, encoder-decoder models, and recurrent networks, alongside a comprehensive review of detection strategies ranging from motion and texture analysis to sensor-based and end-to-end deep learning approaches, the work provides a holistic perspective on the vulnerabilities inherent in current systems and underscores the urgency of developing more robust, adaptable security measures, as well as, clear policies and regulations regarding the use and availability of artificial intelligence tools. |
|---|