The internet is changing rapidly again.  AGIs taking input off of the
internet have to deal with text, audio, images, various symbologies e.g.
mathematics representations, music, different types of protocalled or
standardly formatted information e.g. autocad files, RFC'd communications,
XML markup languages, network transmitted rendering instructions such as in
games and Second Life, etc. etc. but a big one that is happening now rapidly
more and more is VIDEO.  Video in various formats - SWF, mpeg, Windows media
streaming, etc..  Video from what I understand is typically based on a frame
rate, one frame placed on top of another basically sampleable as an ordered
set of images per time unit.  Some video is based on changes between images
depending on compression protocol but it is ultimately a frame rate and a
max visible frame rate to the eye based on monitor refresh frequency.  

Dealing with non-videoed information is very well doable, straight forward
in general; naturally there are still many issues.  Image analysis
technology is there in many ways, audio speech recognition/analysis is
getting pretty good.  But video is looking real nasty.  Lots of resources,
slow, messy, yuk.  Many robots have to do video analysis real-time with
cameras but this usually is very specialized like in speech recognition
where you specify what to look for while it is looking or train it
beforehand.  It seems like there is a whole world of issues to deal with in
Video Mining/Analysis and any pointers in the area especially open source
projects would be interesting to take a look at.

John

-----
This list is sponsored by AGIRI: http://www.agiri.org/email
To unsubscribe or change your options, please go to:
http://v2.listbox.com/member/?member_id=231415&id_secret=28328272-ee06a6

Reply via email to