Abstract. In recent years, Tourism has become more and more Chinese leisure travel choice The research on the smart scenic spot is getting deeper and deeper, but the problem of accurate location l in the natural scenic spot still needs to be solved. Semantic maps contain a wealth of environmental information and can be more efficient for location-aware services, and are attracting more and more attention from researchers at home and abroad. In order to better ensure the travel experience of tourists, the range of scenic spots is too large, and the signal interference is high. Complex terrain in the scenic area, Branch and leaf features Visitors cannot rely on traditional positioning systems to get their current accurate location. It is proposed to construct a navigation semantic map for the perception of scenic space. In the construction process, the operation based on the location perception of the tourists and the surrounding environment and the extraction of the feature information is the key to constructing the semantic map. The general image recognition method is used to obtain the environment image information, and the acquired feature image is recognized to obtain the semantic information in the environment; in order to obtain more feature environment information to better complete the location-aware service task, the GBP descriptor is used. The method divides and stores different semantic regions in the environment, and generates a semantic map with rich semantic information and feature information according to the three-dimensional map model.
Zero-shot object detection aims at incorporating class semantic vectors to realize the detection of (both seen and) unseen classes given an unconstrained test image. In this study, we reveal the core challenges in this research area: how to synthesize robust region features (for unseen objects) that are as intra-class diverse and inter-class separable as the real samples, so that strong unseen object detectors can be trained upon them. To address these challenges, we build a novel zero-shot object detection framework that contains an Intra-class Semantic Diverging component and an Inter-class Structure Preserving component. The former is used to realize the one-to-more mapping to obtain diverse visual features from each class semantic vector, preventing miss-classifying the real unseen objects as image backgrounds. While the latter is used to avoid the synthesized features too scattered to mix up the inter-class and foreground-background relationship. To demonstrate the effectiveness of the proposed approach, comprehensive experiments on PASCAL VOC, COCO, and DIOR datasets are conducted. Notably, our approach achieves the new state-of-the-art performance on PASCAL VOC and COCO and it is the first study to carry out zero-shot object detection in remote sensing imagery.
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.