These datasets are stored in ASCII (text). If using Matlab, they will need to be loaded using the “-ASCII” flag. For example,
load -ASCII speech.mat
speech.mat: “Alexander Graham Bell” sampled at 11,025 Hz
Contains waveform (agb), complex-valued spectrogram (T, F, S)
y.mat: vowel segment
temperature.mat: high temperature and daylight data for Houston
male.mat, female.mat: pitch datasets
classdata1.mat and classdata2.mat: unsupervised learning datasets
Atmospheric signals (“sferics”)
- sferics1.mat (sampling rate: 8012 Hz)
- sferics2.mat (sampling rate: 22050 Hz)
lena.mat: Lena image