Audio analyzer for Speech explorer
Typically reads data from .wav files and outputs json data used by Speech explorer
Feature extraction from wav files can currently be done with mfcc, fft, mel-spectrogram or chroma-stft.
Dimensionality reduction can be performed in 3d or 2d with either t-SNE or PCA.
Additionally audio_concatenator.py can be used to create a binary blob from the audio files used in the crunching process. The blob can be used by the Speech explorer to reduce the amount of network calls the application makes.
