Although current navigation services provide significant benefits to people's mobility, the turn-by-turn instructions they provide are sometimes ineffective. These instructions require people to maintain a high level of attention and cognitive workload while performing distance or angle measurements on their own mental map. To overcome this problem, landmarks have been identified as playing a major role in turnby-turn instructions. This requires the availability of landmarks in navigation databases. Landmarks are commonly selected manually, which involves time-consuming and tedious tasks. Automatic selection of landmarks has recently gained the attention of researchers but currently there are only a few techniques that can select appropriate landmarks. In this article, we present a technique based on a neural network model, where both static and dynamic features are used for selecting landmarks automatically. To train and test this model, two labeling approaches, manual labeling and rule-based labeling, are also discussed. Experiments on the developed technique were conducted and the results show that rule-based labeling has a precision of approximately 90%, which makes the technique suitable and reliable for automatic selection of landmarks.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.