Quote:
Originally Posted by
pik80 
No, there is no way for the artificial intelligence to do this. A computer can't possibly know what the most important words are for a particular file. Fortunately giving a file some tags only takes a few seconds.
I disagree to some extent...
AI can be used to do some things that would be too time consuming to be effectively done manually.
FCP X has some great metadata analysis and generation capabilities -- in addition to making smart collections and tagging your folder structure and camera/file metadata, FCP X can scan and analyze AV files for:
1) People
-- 1 person
-- 2 people
-- a group of more than 2 people
2) shot type
-- closeup
-- medium
-- wide
3) color balance
4) stabilization and rolling shutter
5) audio problems
-- background hum
The results of these analyses are placed in smart collections and can optionally be corrected automatically,
Now the clip exists only once -- but it can appear in many smart collections based on file structure keywords, camera metadata, user assigned keywords (tags)...
So, for example, you could set up a smart collection that contained any 720p clips, from your Sony camera, which contain a medium shot of 2 people, needs stabilization and has camera hum...
Then any clip you ingest will be automatically tested and placed in this (or any other) smart collection that matches the criteria,
This is easy to use and superior to OSX smart folders.
I suspect we will soon see this in OSX ML,,, then later a compatible, lighter version for iOS.