Context-based Media Geotagging of Personal Photos

This paper addresses the problem of automatic geotagging of media within the context of a personal media collection. In contrast with textual and visual methods which tackle the same problem we ap- proach it focusing on analysis of contextual information. An event as a context aggregator plays the central role in our approach. The proposed method automatically estimates geographical coordinates (latitude and longitude) within the temporal boundaries of events computed from a personal media collection. Proposed framework interpolates or extrapolates GPS information rely on geoannotated media entities from the collection. The process of interpolation is automatically performed by the framework based on temporal distances between samples in combination with using free on-line navigation service. All this leads to a new cost efficient and intelli- gible event-centered way to enrich the collection with geographical information. Experimental results show that we are able to assign geographical coordinates for 83% of images within an error of 5 km.