Discriminative pronunciation modeling for dialectal speech recognition

   Abstract