Author_Institution :
Univ. of Washington, Seattle, WA, USA
Abstract :
Summary form only given. Collectively, we take upwards of a trillion photos each year. These images together comprise a nearly complete visual record of the world´s people, places, things and events. However, this record is massively disorganized, unlabeled, and untapped. This talk explores ways of transforming this massive, unorganized photo collection into reconstructions and visualizations of the world´s sites, cities, and people. After a brief recap of our work on Photosynth and reconstructing Rome, I will present new work on matching, reconstructing, and browsing people from large photo collections.
Keywords :
data visualisation; image matching; image reconstruction; image retrieval; visual databases; Photosynth; disorganized record; image reconstruction; large photo collections; people browsing; people matching; people reconstruction; unlabeled record; unorganized photo collection; untapped record; visual record;
