There are other ways though e.g. if features like formants produced by the model can be differentiated with respect to vocal tract parameters, the latter could

There are other ways though e.g. if features like formants produced by the model can be differentiated with respect to vocal tract parameters, the latter could be estimated based on real data maybe