On Adobe’s Lightroom 4 Beta forum, one of the most contentious topics has been the omission of automated facial recognition or facial detection feature. If you don’t get the distinction, one is the computer trying to figure out who someone is by comparison with existing pictures, the other is it doing only the first step in that process and finding faces so you can identify them manually.

While I consider it as important as of those two features, I agree with it receiving a lower priority than books and geotagging. Apart from the likely demands on the computer, the trouble is that if face recognition data is ever added, it will need proper provision to keep that information private if the user wishes. So rather like keywords can be marked not to export, data about people needs to be fully controlled – with the default being that it isn’t written to exported files.

I am certainly in favour of recording who is in a photo in a more structured way – using the existing Person Shown field but in a keywords-style panel and with image areas defined. I’m not against face detection or automated recognition in principle, just not if the price is failing to deal with existing weaknesses in Lightroom’s DAM features.

What do you think?