DeepFake Audio Detection using Machine Learning
author, Mrs .Chandana V S · INTERANTIONAL JOURNAL OF SCIENTIFIC RESEARCH IN ENGINEERING AND MANAGEMENT · 2025
Abstract Our system listens deeply to what makes human speech uniquely human. It picks up on the natural patterns and subtle imperfections that even the best AI voices haven't quite mastered yet. Think of it as teaching a computer to hear the way your ear instinctively knows when something sounds "off" about a voice. We've wrapped this technology in a simple website where anyone can upload a recording and get a quick answer: real person or clever fake? With accuracy that catches 93% of imposters, it's like having a trusted friend who can spot when something doesn't sound quite right. This matters because voice is personal. When someone calls claiming to be your child in trouble or your bank with an urgent message, you deserve to know it's really them. We're working to protect that fundamental human connection in our increasingly digital world.