Voice forgery detection in voice-based communication systems is often performed using offline methods,
which are slow and ineffective during live calls. The increasing use of AI-generated voice cloning makes
it difficult to distinguish real voices from fake ones in real time. This project proposes “An Adaptive Real-
Time Voice Forgery Detection System using Hybrid Intelligence Techniques” to identify forged voices
during ongoing conversations. The system analyzes live voice signals by applying hybrid intelligence
methods that combine deep learning and rule-based analysis. It provides real-time alerts, improves
detection accuracy, and ensures privacy, offering an efficient and reliable solution for preventing voicebased
fraud.
