4 ms·
> But once solved, we should try to reverse engineer the solution and optimize. This is a great point. Does anybody know of any papers that explore non-neural
by echelon 5y ago
> But once solved, we should try to reverse engineer the solution and optimize.
This is a great point.
Does anybody know of any papers that explore non-neural text to speech / voice synthesis that achieve better than parametric quality?