Building a Machine Learning Model for Answering Machine Detection

This project involves building an answering machine detection system using a trained machine learning model, specifically a Gaussian Naive Bayes classifier, which achieved 96% accuracy. The system uses audio samples of beeps and speech to train the model, and then uses the trained model to detect when an answering machine is on a voice call. When a beep is detected, the system sends a message saying "Answering Machine Detected" into the call. The project also involves building a client application that connects to a websocket, observes when a beep is detected, and sends a TTS into the call when a voicemail is detected. The system uses Python libraries such as Scikit-learn, Librosa, and Matplotlib for machine learning and audio processing tasks.