This was circulated today in our lab and I thought it might be of interest to those of us who are doing work in crowdsourcing.
It is basically a camera that instead of producing an image, this prototype outputs a text description of the scene using Amazon’s Mechanical Turk API. I thought the concept and the results might be of interest to the Oxford crowd. It also reminded me of Khaled mini crowdsourcing exercise during the hackfest in Nottingham which is now followed up to be deployed in a bigger scale.

http://mattrichardson.com/Descriptive-Camera/?src=twitter