skip to content

All projects: Gel, Jobs, Gootodo, Games, Uncle Mark, Goovite, Blog, Bit Literacy

Problems with finding smiles in photos

Here's another example of technology trying to solve a problem that would be handled with a few minutes of bit literacy training.

For those camera buyers who haven't read the managing photos chapter in Bit Literacy (summary: take tons of shots up front, then filter later to a very few good ones), camera companies are touting a shiny new technology - an automatic smile-finder, which will identify "good" photos for you. (Darn those moments when we actually have to think for ourselves!)

As Daniel Rutter points out in this post, the SmileCheck really does find toothy grins:

smilecheckafter500.jpg

Mmm, good photo!


9 Comments:

Zephyr — Jan 15, '08 — 2:48 PM

I'm sentimental for times when smiling in photos was not required (pre-1950s?)

Carlos Gomez — Jan 16, '08 — 1:08 PM

I suppose that it might be vaguely useful in a point and shoot camera for those people taking snapshots where they want people smiling.

But I agree with Mark. Using a digital camera (and a suitably large memory card) just get snap happy and snap away. Some of your best shots can be of people when they are not smiling, but are caught in a more natural state.

As Zephyr states, smiling for a photo really doesn't need to be a requirement.

ABD — Jan 16, '08 — 2:54 PM

This is a very concrete example of how embarrassingly useless sophisticated technology workarounds for very simple problems can be. Glad I bought your book--maybe you should also be marketing it as a digital camera accessory?

Mark Hurst — Jan 16, '08 — 3:00 PM

I'd be happy to - just tell the camera manus to get in touch with me :)

Michelle — Jan 16, '08 — 6:10 PM

This reminds me of the HP cameras that are supposed to take 10 pounds off the photo subject by subtly "squeezing" that part of the image. Something that can be accomodated with a little human interaction (posing the subject skilfully).

Jonathan — Jan 17, '08 — 3:49 AM

I think you're missing the point of the smile finder. As I understand it, the idea is that you can put the camera on a tripod and then set the self-timer to automatically take the picture at the time when all subjects are smiling, rather than an arbitrary 10 seconds after you start the self-timer.

Mark Hurst — Jan 17, '08 — 9:27 AM

That's a related feature, FaceTime... the drawback there is - what happens if one person isn't smiley enough for the camera - how long do people have to stand there waiting?

Reminds me of what HAL might have said: "I'm sorry, Dave, I can't take the picture. John still looks sad."

Julia — Jan 17, '08 — 9:27 AM

@ Michelle: I'll be glad when a camera can *actually* take 10 pounds off a photo subject. That'll be the day! ;)

Patrick Rogers — Jan 18, '08 — 11:49 AM

In all fairness, this technology isn't being developed with digital cameras in mind. Facial expression software is really about things like improving the computer/human interaction, recognizing drivers who are becoming drowsy and picking out potential terrorists when they are boarding a plane. In fact, it could conceivably be very useful in computer-aided usability studies. When added with eye-tracking, process-tracing, etc a facial mood recognition algorithm could include data on user's annoyance factor.

That said, this is definitely an example of companies adding unnecessarily complicated bells and whistles so they can add a new bullet point on their packaging, with no consideration of the user's experience.




All Projects from Good Experience

Gel Conference
Our annual get-together in New York
Jobs Board
Post or find a job
Gootodo
The world's best todo list
Good Experience Games
The best games online
Uncle Mark Gift Guide
The guide to technology and life
Goovite
Easy event invites
Good Experience Blog & Newsletter
Mark Hurst explores good experience

"...the Elements of Style for the digital age."
- Seth Godin
Bit Literacy, the book by Mark Hurst, shows how to solve email and info overload.