This is an Artificial Intelligence designed to identify instruments from audio samples. The data from the audio samples will be analyzed and presented in different formats to the AI, such as raw wavelength data, a Fourier Transform, and a spectrogram.
Use pip, python's default package installer, to install several dependencies. In the command prompt, run these commands:
pip install numpy
pip install scipy
pip install librosa
pip install matplotlib
pip install tensorflow
pip install scikit-learn
pip install opencv-python
Then, you can run main.py using either the command prompt or any Python development environment.
The finished AI will be used to test our main project, the Octopi. Since AIs take a long time to train, it is a good test for the computing power of the network of Raspberry Pis compared to a single Pi or a personal computer.