brolsergii/ErrViz
ErrViz is a tool for linguists which helps them to find and visualise the errors of automatic voice recognition systems. It uses STM (Refrences), CTM (Hypothesis) and WAV file formats as the input data.
Discovered public repositories for brolsergii in the GitHub catalog.
ErrViz is a tool for linguists which helps them to find and visualise the errors of automatic voice recognition systems. It uses STM (Refrences), CTM (Hypothesis) and WAV file formats as the input data.