How to categorize Friends and Family Face Annotation Keywords

[rebus]

Hi folks,

I'm looking at switching from Lightroom to iMatch to organize and catalog my photos, but I've got one thing I can't figure out how to do.

In Lightroom, my keyword hierarchy like this:

+
|
+- Events
|  |
|  + ...
|
+- People
|  |
|  +- Family
|  |  |
|  |  +- Bob
|  |  +- Mary
|  |
|  +- Friends
|     |
|     +- Jack
|     +- Jill
|
+...

Each of the people (Bob, Mary, Jack & Jill) are Person keywords and I use them when tagging face regions in photos. This structure lets me easily find photos that contain family members or friends.

As I understand face annotations in iMatch, they have to be a flat hierarchy under a single keyword (i.e. WHO|Bob, WHO|Mary, etc).

• Is there a way to have a more hierarchical structure to face annotation keywords (e.g. WHO|Family|Bob, WHO|Friends|Jill)?
• If not, is there another way to achieve something similar in iMatch? It seems like data driven categories should the answer, but I've not been able to get one working for "Friends"  I was trying an inverse filter to exclude family, but then the category also includes other keywords that the pictures contain.

Thanks for any tips or suggestions!

[Mario]

By definition, a XMP face region has a 'tag', which is a single word or text. No relations to keywords, no hierarchy or anything.

IMatch allows you to integrate XMP face region tags into your hierarchical keyword structure using the options provided under Edit > Preferences > Metadata2 : Face tags as keywords.

Here you tell IMatch that the tag you enter for a face should be converted into a hierarchical keyword, and placed under the root keyword you configure there. For example the keyword WHO|person.

If you now add the tag "Susan" to a face annotation, IMatch creates the corresponding keyword WHO|person|Susan for it.
You can also add additional levels, e.g. by specifying the tag as "Family|Miller|Susan". This creates the keyword WHO|person|Family|Miller|Susan which is probably what you have in mind.

Tip: You can create categories under the face root category before you start tagging persons. This way you can directly pick the person name from the tag lookup dialog in the Viewer.
If you have tons of images with existing "flat" fade regions (name only), IMatch will use your Universal Thesaurus during the import to map the flat keywords imported from the face tag into your @Keyword hierarchy. This allows you to automatically transform the tag "Tom" to  WHO|person|Family|Miller|Tom if you want.

Please note that the XMP face region standard only defines the region tag as  "word", without any special meaning. Different applications may interpret it in different ways. Adobe often re-defines their own XMP standard to whatever the currently need for one of their applications, not necessarily caring for other applications or long-term compatibility. Adobe writes their applications with an "Adobe-only" approach, assuming the only products you use are from Adobe.

IMatch is very flexible when it comes to handling metadata and can deal with almost everything that is out there. On the other hand, IMatch greatly favors established metadata standards and tries to create metadata that is accessible and compatible with the widest possible range of applications, workflows and tool sets.

While using | as a hierarchy separator for face tags is not per-sé disallowed, it might cause issues with other software you use (now or in the future) so make sure to test it properly. IMatch has no problem with it and interprets the | as a standard hierarchy separator for keywords when mapping the XMP face region tag into keywords. Other applications may see this differently. Not that there are many applications out there which support XMP face regions. Not all applications are as flexible of configurable as IMatch.

A side note. often users who leave the Adobe Cocoon for some reason suddenly realize that many things they are used to only work within an Adobe product line or even only one Adobe product / version.

[rebus]

Mario,

Thanks for the detailed reply! It hadn't occurred to me to use a hierarchical tag for the face region - that's exactly what I was looking for.

I appreciate the warning about standards support; right now I'm expecting to use iMatch/iMatch Anywhere as my main interface for finding photos so I'm not too worried about how other sofware supports a hierarchy here. In fact, the emphasis iMatch places on standards-compliance is a big point in its favor.