What layer / library is used for the actual speech recognition part? I couldn't find it now. ALSO: commits actually go back to 2016, so it's been worked on fo

What layer / library is used for the actual speech recognition part? I couldn't find it now. ALSO: commits actually go back to 2016, so it's been worked on for a while.