A many-to-one phone mapping approach for cross-lingual speech recognition

This paper presents a novel method for acoustic modeling of an under-resourced language by “mapping” from acoustic models of well-resourced languages. The proposed method can be considered as a “many-to-one mapping” method where one speech unit in the target language is built as a linear combination of the source speech unit models and hence we can explicitly observe the relationship of the source language and target language phones. Experiments conducted on the WSJO task show that the proposed method can obtain a much higher performance over the other phone mapping methods.