+2
Shelved

Aprimo AI feature - Auto tag an 'orientation' field

katie.tierney 12 months ago in Digital Asset Management updated by petra.tant 11 months ago 5

For Nord Anglia, we would like to see if the Aprimo AI functionality will be able to auto tag an 'Orientation' field for images. This is helpful for users searching for the correct image to use in their marketing material, ideally, they would want to locate the best image to use initially without having to image crop, so this orientation field would be helpful. If users don't have to populate this when they are uploading, then it is a time saving / click saving benefit!

Fields / Attributes AI
Shelved

Because the use case can be handled with this reference, we have currently no plans to prioritize this as a standard AI solution. But i will leave it up for voting.

A small note: you can already surface this info via a (complex) reference:

<ref:catch ex="System.Exception" out="">
<ref:record file="master" out="info" xpath="/metadata/movie/videoWidth" store="@width"/>
<ref:record file="master" out="info" xpath="/metadata/movie/videoHeight" store="@height"/>
<ref:record file="master" out="info" xpath="/metadata/raster/pixelsWidth" store="@width" onVariable="IsEmpty(@width)" />
<ref:record file="master" out="info" xpath="/metadata/raster/pixelsHeight" store="@height" onVariable="IsEmpty(@height)"/>
<ref:eval expression="@width/@height" store="@aspectRatio"/>
<ref:compare value1="1" value2="@aspectRatio" store="@isSquare"/>
<ref:text out="Square" onVariable="IsNotZero(@isSquare)"/>
<ref:text onVariable="IsZero(@isSquare)">
<ref:eval expression="@aspectRatio - 1" store="@aspectRatioMinusOne"/>
<ref:text out="@aspectRatioMinusOne" left="1" store="@firstCharacter"/>
<ref:compare value1="-" value2="@firstCharacter" store="@isPortrait"/>
<ref:text out="Portrait" onVariable="IsNotZero(@isPortrait)"/>
<ref:text out="Landscape" onVariable="IsZero(@isPortrait)"/>
</ref:text>
</ref:catch>

For us, the value doesn't need to be in its own distinct field, but would be great to have captured automatically, and then be able to filter on it.