With the tremendous explosion of community photos, how can we present a coherent visual model for a large-scale community photo collection? In this project, we aim to build a photo space explorer for effective image browsing, retrieval, and annotation on top of our existing, distributed, community storage mechanism. The key is to identify object and feature points relations among community photos in an effective and scalable way, and to distribute the storage through an ensemble of mobile devices. The photo space explorer also enables users to navigate community photo collections using a 3-D browser.