[09:58:39] jayvdb: So the 3 buckets Im adding are: "Type", "Content" and "Location" [09:59:06] What bucket would the "Taken with Canon XYZ" and "Created by ..." categories do into ? [09:59:14] type [09:59:51] So,e of the "Created by ..." are similar to content - because if something is Inkscape it would *probably* be a Graphic, or Chemtool => it's a chemical compound, etc [09:59:58] Some * [10:00:15] Graphic is also type [10:00:27] Category:Graphic is a type o.O ? [10:00:40] a picture of a person can be a graphic or a photograph [10:00:44] How come ? Doesn't it tell the information that's in the image ? [10:00:46] Hmmm [10:00:57] Ok [10:01:08] a drawing or an realistic capture of a slice of time [10:01:20] Black and white photograph ? [10:01:20] a painting is a type [10:01:24] type [10:01:30] Alright [10:01:37] content is computer vision [10:01:53] type & location are mostly metadata [10:02:14] So, based on the current classification of buckets - It would seem only faces and barcodes are in "Content" [10:02:23] detecting a shrub is content [10:02:45] What about: Detecting if an image is a screenshot ? [10:02:52] Would that also be Type ? [10:02:54] that is content [10:03:08] Hmm, ok [10:03:17] I think I should make a list and run it by you to avoid confusion [10:03:26] * AbdealiJK makes one now [10:03:43] it is black and white, that is the type. it if is 640x480, that is type [10:05:40] if it is 32x32, and therefore an icon, that is type. transparency is type. but if you detect that it is a white background that should be converted to transparency, that is content [10:06:34] I see, ok - that does make intuitive sense [10:10:56] jayvdb: https://etherpad.wikimedia.org/p/catimages_buckets [10:29:57] "Scanned with IT8 Strip" is content, why ? [10:40:28] I was of the impression it detects something abotu the image and hence would be in content [10:43:30] jayvdb: The IT8 Strip is indeed debatable [10:43:52] You can decide either Type or Content and I can put that