🔊 Sound Classification

Simple sound classification in Python using a Convolutional Neural Network (CNN).

🚀 How it works

Audio files are converted into NumPy arrays using Mel-spectrogram conversion.
This process turns the raw audio data into a spectrogram image, capturing frequency vs time.
The spectrogram is then used as input for a CNN model, which learns to classify different types of sounds.

🧰 Features

• Convert audio files into spectrograms using the MEL scale
• Feed image-like data into a CNN for training and prediction
• Easily customizable for different sound classification tasks

🛠 Requirements

• Python 3.8+
• TensorFlow / Keras
• Librosa
• NumPy
• Matplotlib (optional for spectrogram visualization)

Running

To train the model, use the train.py script:

python train.py

To use the model use the test.py script:

python test.py

Training