Currently, the tools are available for Mac Os X and Linux, and have been tested on OS X El-Capitan and Ubuntu 15.04. Windows versions of the tools are in development.
Currently, the tools are not yet available for Windows. However, Windows users can download a Linux virtual machine through VirtualBox. You can create your own machine set up to your specifications, or you can download a pre-existing machine. We recommend the Berkeley Phonetics Machine, which comes with several phonetic and analytical tools pre-installed, including Praat, wavesurfer, and sox.
Currently, these tools work with recordings consisting of a single word. AutoVowelDuration assumes that the word contains a single vowel, preceeded or and followed by one or more consonants. AutoVOT can make predictions for aspirated or unaspirated word-initial stops; in addition to these, DeepVOT can also handle prevoiced stops.
You can find more information on our lab's git repository: ML.Speech Research Lab