home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!stanford.edu!unix!unix.sri.com!laws
- From: laws@ai.sri.com (Kenneth I. Laws)
- Newsgroups: comp.ai.neural-nets
- Subject: Re: Kahaner Report: Facial classification by neural nets
- Message-ID: <LAWS.93Jan10211159@sunset.ai.sri.com>
- Date: 11 Jan 93 05:11:59 GMT
- References: <29152@optima.cs.arizona.edu> <1993Jan6.135012.13411@bsu-ucs>
- Sender: news@unix.SRI.COM
- Followup-To: comp.ai.neural-nets
- Organization: Computists International
- Lines: 16
- In-reply-to: 01jmbrown@leo.bsuvc.bsu.edu's message of 6 Jan 93 18:50:12 GMT
-
- Someone may already have mentioned this ...
-
- Before I can get excited about the gender-recognition announcement,
- I have to know a bit more about how the images were obtained or
- normalized. In the worst case, suppose that the training and
- test images were digitized from two class photographs -- a men's
- class and a women's class. It might then be trivial to distinguish
- them on the basis of mean gray level. The problem becomes easier
- as the resolution approaches one pixel per face.
-
- -- Ken
-
- --
-
- Dr. Kenneth I. Laws; (415) 493-7390; laws@ai.sri.com.
- Ask about my AI/IS/CS newsletter, The Computists' Communique.
-