Articulatory and Spectrum Features Integration Using Generalized Distillation Framework