home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!charon.amdahl.com!pacbell.com!ames!saimiri.primate.wisc.edu!zaphod.mps.ohio-state.edu!uwm.edu!ogicse!das-news.harvard.edu!husc-news.harvard.edu!husc8.harvard.edu!mlevin
- From: mlevin@husc8.harvard.edu (Michael Levin)
- Newsgroups: comp.ai.neural-nets
- Subject: need info on using neural nets to pick out visual features
- Message-ID: <1992Nov6.151006.17167@husc3.harvard.edu>
- Date: 6 Nov 92 20:10:05 GMT
- Article-I.D.: husc3.1992Nov6.151006.17167
- Distribution: usa
- Organization: Harvard University Science Center
- Lines: 15
- Nntp-Posting-Host: husc8.harvard.edu
-
-
- I would like to investigate the feasibility of using neural nets
- to pick out certain features in photographs taken through a
- microscope. The way it is done now, is to take a photo, and then go
- over the desired features by hand (using a track-ball device). This is
- a very difficult thing to formalize, but people can do it with lots of
- practice. I want to look into the possibility of having the neural net
- learn to do it. We have a database of thousands of sample cases built
- up over the years (and already scanned into a computer). I'd like some
- info on references to
- 1) books or papers that describe the kind of neural networks used for picking
- out visual features (pattern recognition, classification, etc.)
- 2) descriptions of the algorithm used to feed an image to a network, etc.
- 3) any existing network software packages in the public domain, etc.
- Please reply to mlevin@husc8.harvard.edu if you have any info.
-