Foreign language pronunciation interactive speech recognition learning method having computer data base vocabulary and similar/different phonetic parts recognition/vocabulary list comparison.

摘要

The interactive learning speech recognition technique has a data base vocabulary. During repetition of words being learnt, incorrect pronunciation is separated from correct pronunciation word parts. Phrases are recognized from the vocabulary, and the nearest incorrect word found with respect to the reference phrase. The speech recognition interactive learning technique has a vocabulary base containing different words. From these words a number of reference phrases are determined. During speech recognition the nearest phonetic words are found. These words are then compared to the active vocabulary to determine which words are phonetically the nearest and whether the phrase is correct or not.