Framework for cross-language automatic phonetic segmentation