From image retrieval to image classification, all research shares one common requirement: a good image database to test or train the algorithms. Such a database is often difficult to find and must therefore be created by the people who need it. As a consequence, the database is influenced by the researchers’ needs. Often it does not contain a sufficiently large number of images.
In order to create a large database of images, we set up a project that allowed us to gather a collection of more than 33000 photographs with keywords and tags from all over the world. This project was part of the “All Photographers Now! The rapid mutation of amateur photography in the digital age” exhibit at the Musee de l’Elysee in Lausanne. The “Flux”, as it was called, gave all photographers, professional or amateur, the opportunity to have their pictures shown in the museum. Everyone could upload pictures on a website (www.allphotographersnow.ch). We required that some simple fields (Title, Subject, Location, Date) were filled in. Keywords were optional. The pictures were then projected at the museum. A webcam “snapshot” (i.e. an “exhibition view”) was taken and sent back to the photographers via email to show how and when their picture was displayed at the museum (Figure 1).
We collected the information in a MySQL database along with the original picture. The latter might also contain some EXIF information, if the editing process didn’t remove it. Information on photographers’ location was also gathered through an optional field and IP address. The backend of this system consisted of a web server running PHP/MySQL to handle the picture uploads. AJAX was used for the moderation page, to manage the slideshow at the museum, and to synchronize it with the webcam.
During the 14 weeks of the exhibition, we collected more than 33000 JPEG pictures, with tags and keywords. These pictures come from 133 countries and were taken by 9042 different photographers. This database can be used for non-commercial research at EPFL.
Project Period :
October, 2006 – May, 2007.
Funding source :