Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Creative Labs (CREAF) -- Ignore unavailable to you. Want to Upgrade?

To: Seth49 who wrote (5914)	11/19/1997 9:51:00 AM
From: baney	Read Replies (1) \| Respond to of 13925

27 13/16 X 28 last 27 7/8 + 13/16

To: Seth49 who wrote (5914)	11/19/1997 11:33:00 AM
From: Douglas Webb	Respond to of 13925

I did some consulting programming work with Lernout & Hauspie's speech recognition software a year ago. The concept was that if you create a dictionary of words which includes their context, then a voice-independent speech recognition program would have a high success rate. The problem is that you can't do free-form text recognition. The dictionary is setup so that each word can only be followed by a specified set of other words; if the user says something that's not in the pre-programmed vocabulary, the recognition fails. For a command-control system with a well-defined and limited vocabulary the system will work great, but that's a niche market at best. I personally feel that a graphical iconic touchscreen is much easier to use than this speech recognition product was. On the technical side, the software library was a pain to develop with too. It was all 16-bit windows 3.1 code, and not very robust. (I somehow managed to wrap a multithreaded java program around it, but that was definitely prototype-quality code.) Over the past year they may have rewritten the library in 32-bit c++, of course. Doug.