The state of the art in content-based image retrieval in P2P networks
ICIMCS '10 Proceedings of the Second International Conference on Internet Multimedia Computing and Service
Hi-index | 0.00 |
Hunt the Cluster (HTC) is an interactive image browser for small to medium-sized P2P networks that uses methods that are likely to scale also to larger net- works. In Peer-to-Peer (P2P) networks, computers with equal rights form a logical (overlay) network in order to provide a common service that lies beyond the capac- ity of each single participant. Current research on re- trieval in P2P systems focuses largely on keyword-based retrieval and other weakly interactive query paradigms. In contrast, interaction is at the center of our contri- bution. We demo a Bayesian, PicHunter-like image browser that helps the user in finding images in distributed col- lections. Such browsers operate by presenting sequences of thumbnail-based collection overviews to the user, at each step collecting user feedback that is used to update a Bayesian model. Our approach is scalable in that the state of the Bayesian model is maintained locally in the browsing peer and only a few thumbnails are requested from the network at each step. Each query step is thus done in a short timeframe. At the conference we will present one peer that is connected to a P2P network consisting of 50 peers, let- ting visitors try the browsing functionality of our sys- tem.