No abstract

In this study, we explore the use of deep convolutional neural networks (DCNNs) in visual place classification for robotic mapping and localization. An open question is how to partition the robot's workspace into places to maximize the performance (e.g., accuracy, precision, recall) of potential DCNN classifiers. This is a chicken and egg problem: If we had a welltrained DCNN classifier, it is rather easy to partition the robot's workspace into places, but the training of a DCNN classifier requires a set of pre-defined place classes. In this study, we address this problem and present several strategies for unsupervised discovery of place classes ("time cue," "location cue," "timeappearance cue," and "location-appearance cue"). We also evaluate the efficacy of the proposed methods using the publicly available University of Michigan North Campus Long-Term (NCLT) Dataset.

Track processing is the foundation of radar multi-target tracking, and the processing performance for jamming has particular research significance when it comes to protecting high-value targets. At present, passive jamming using a modulated metasurface exhibits a fast response and a flexible operation mode. However, most research in this area has been carried out at the radar signal processing level and less at the data processing level. In this paper, a range of false target track deception method based on a phase-switched screen (PSS) is proposed, and the relationship between the matched filtering output, radar detection, and track processing is derived. This method uses PSS to generate multiple false targets with controlled spatial distribution and magnitude, which can form high-fidelity false tracking tracks. The number of false tracking tracks can be flexibly altered by controlling the modulation parameters. The simulation results validate the effectiveness of the proposed method.