Fake COVID-19 videos detector based on frames and audio watermarking

Area of Science:

• Computer Science
• Information Security
• Digital Forensics

Background:

• Advanced video editing technology necessitates robust methods for verifying video authenticity.
• Malicious alterations to sensitive video content (surveillance, medical) can lead to flawed decision-making.
• Ensuring the integrity of video information is crucial in an increasingly digital society.

Purpose of the Study:

• To develop and evaluate a novel fake video detection system.
• To enhance the security and reliability of video information.
• To demonstrate the system's efficacy on real-world data, such as COVID-19 news.

Main Methods:

• A fake video detector was designed by integrating audio and frame watermarking techniques.
• The system was tested for its ability to detect unauthorized modifications.
• Experimental validation was performed using specific datasets, including COVID-19 video news.

Main Results:

• The developed system demonstrated good robustness against potential tampering.
• Effective integrity verification capabilities were confirmed through experimental testing.
• The detector proved successful in securing sensitive video content.

Conclusions:

• Combining audio and frame watermarking offers a viable solution for fake video detection.
• The proposed method ensures the integrity and reliability of video information.
• This approach is particularly valuable for securing critical video data in various applications.