Speaker Independent Speech Recognition with Neural Networks and Speech Knowledge