This user manual briefly describes the main features of the speech
processing toolkit SPro 3.3.1. Basically, SPro was designed
for variable spectral resolution analysis of the speech signal (see
section for more details on this topic). It provides a
library for the standard (frame-level) analysis techniques and for the
signal and feature I/O functions. Classical feature extraction
algorithms are also provided as runtimes, developped using the
library.
In section , the file formats used by SPro are
described. In section
, the basic feature
extraction tools are described. Finally, an overview of the functions
provided by the library is given in section
.