home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!mcsun!uknet!cam-eng!ajr
- From: ajr@eng.cam.ac.uk (Tony Robinson)
- Newsgroups: comp.ai.neural-nets
- Subject: Re: Temporal networks
- Message-ID: <1992Aug13.141932.4315@eng.cam.ac.uk>
- Date: 13 Aug 92 14:19:32 GMT
- References: <1992Aug13.125946.3197@hubcap.clemson.edu>
- Sender: ajr@eng.cam.ac.uk (Tony Robinson)
- Organization: Cambridge University Engineering Department, UK
- Lines: 13
- Nntp-Posting-Host: dsl.eng.cam.ac.uk
-
- In comp.ai.neural-nets bdbryan@eng.clemson.edu writes:
- >I have recently been doing some work with speech recognition, and have used a
- >few network configurations in my search for "the perfect network architecture".
- >I wanted to find out from people in this group what some of the best networks
- >for sequence recognition are?
-
- Sequence recognition or speech recognition? I'll design you a HMM-net for
- sequence recognition that will be optimal (according to my definition :-).
- For speech recognition, you (and everyone else) are welcome to look at the
- reports on recurrent nets for phone/word recognition available by anonymous
- ftp from svr-ftp.eng.cam.ac.uk in directory "reports".
-
- Tony [Robinson]
-