Publication

Combining Acoustic Data Driven G2P and Letter-to-Sound Rules for Under Resource Lexicon Generation