Human or Not - Can You Really Detect the Fake Voices?

Explore the challenges of detecting AI-synthesized speech in a 25-minute Black Hat conference talk. Delve into the importance of voice as a medium for information transfer and trust-building, and examine how deep learning technologies have enabled voice synthesis and cloning. Analyze the effectiveness of proposed AI-synthesized speech detection approaches in real-world scenarios compared to laboratory environments. Learn from presenters Xin Liu, Yuan Tan, Rui Chong, Xiaokang Zhou, Mingyuan Zhan, and Qingguo Zhou as they discuss the complexities of distinguishing between human and artificially generated voices. Access the full abstract and presentation materials for a comprehensive understanding of this critical cybersecurity topic.