RECOGNITION ARCHITECTURE FOR GENERATING ASIAN CHARACTERS

摘要

An architecture that corrects incorrect recognition results in Asian language speech recognition systems. In response to receiving the speech input, a spelling mode may be activated to correct erroneous spelling of the recognition result or to generate a new word. Correction can be achieved using voice and / or manual selection and input. This architecture facilitates correcting more than once, as in conventional systems. Using the spelling mode, the corrected word is corrected as a unit and treated as a word. The spelling mode applies to at least the languages of the Asian continent (such as Simplified Chinese, Traditional Chinese, and / or other Asian languages such as Japanese).;Speech recognition, spelling mode, engine dictionary, acoustic model, language model