User Study of Automatic Photo Classifier by Color and Timestamp

We developed a new web-based intelligent photo browsing system which helps to find desired photos more efficiently from unfamiliar photo collections. We conducted a user study to assess the effectiveness of the developed photo browser (automatic photo classifier by color and timestamp, APC-CT) compared to ones with other clustering methods. The user task adopted here was to find some target photographs indicated by the experimenter from somebody else’s photo collections. The results show that APC-CT makes it faster to find the target photographs. Also, we confirmed that clustering using timestamp makes photo browsing more efficient even for unfamiliar photo collections.