Speech Recognition
Voice Synthesis
For alphabetic languages, speech recognition so far is NOT sucessful.
For ideographic languages such as Chinese, speech recognition is even harder,
IBM and LUCENT tried, but, so far NO
acceptable results yet.
Chinese has about 97% characters having
same / similar pronunciation
beside hundreds of dialects.
Here are some recent reports:
Bell Lab
Handwriting Recogntion
Xerox has the patent of
UNISTROKE TECHNOLOGY
WowOh.com