The national Speech Corpus for ASR of the year 1996, also funded by the National 863 High-Tech Program, was mainly of reading style, considering the phonetic balance between segments.

• 1996年863语音识别数据库以朗读语体为主，考虑了语音的音段平衡。