US7415411B2 - Method and Apparatus for Generating Acoustic Models for Speaker Independent Speech Recognition of Foreign Words Uttered by Non-Native Speakers