This project implements the MelScalogram class, a tool for audio signal analysis that combines continuous wavelet transform techniques with the Mel scale for data visualization and processing. It is particularly useful for applications in signal processing and machine learning, providing methods to visualize and pre-process data for convolutional neural networks.
- Generation of scalograms using real and complex wavelets.
- Conversion of frequencies to the Mel scale.
- Various subsampling methods (mean, maximum, minimum, uniform).
To run MelScalogram, you will need Python 3.10.12 and some specific libraries. Install them using the following command:
pip install -r requirements.txt