Semantic navigation of large scale geo-referenced 3D scenes and virtual worlds
Christos P. Yiakoumettis · OpenGrey (Institut de l'Information Scientifique et Technique) · 2012
The current technological evolutions introduce 3D geo-informatics to their digital age, enabling new potential applications in the field of virtual tourism, recreation, entertainment and cultural heritage. Virtual 3D worlds and geo-referenced scenes are used even more in simulations of physical disasters or in evacuating and military scenarios. It is argued that 3D information provides the natural way of navigation. However, personalization is a key aspect in a navigation system, since a route that incorporates user preferences is ultimately more suitable than a route which simply provides the shortest distance or travel time. An efficient personalized route planning architecture is based on geometrical criteria and on human factors regarding user's preferences in the 3D itinerary. In this work a multi-layer integrated framework is introduced for efficient, user-centric navigation of 3D worlds. Usually, user's preferences are expressed as a set of weights that regulate the degree of importance of the scene semantic metadata on the route selection process. These weights, however, are defined by the users, setting the complexity on the user's side, which makes personalization an arduous task. In this work, an alternative approach is proposed in which metadata weights are estimated implicitly and transparently to users, transferring the complexity to the system side. This is achieved by introducing a relevance feedback online learning strategy which automatically adjusts metadata weights by exploiting information fed back to the system about the relevance of user's preferences judgments given in a form of pairwise comparisons. Practically implementing a relevance feedback algorithm presents the limitation that several pairwise comparisons (samples) are required to converge to a set of reliable metadata weights. For this reason, in this work a weight rectification strategy is proposed which improves weight estimation by exploiting metadata interrelations defined through an ontology. In the sequel, a genetic optimization algorithm is incorporated to select the user's most preferred routes based on a multi-criteria optimization approach. To increase the degree of personalization in 3D navigation, an efficient algorithm is also introduced for estimating 3D trajectories around objects of interest by merging best selected 2D projected views that contain faces which are mostly preferred by users. Simulations and comparisons have been conducted with other approaches either in the field of online learning or route selection using objective metrics in terms of precision and recall values.