Nowadays automatic methods based on artificial intelligence are rapidly growing. In the paper, a problem of automatic target recognition in synthetic aperture radar images is described. It is demonstrated, that two different machine learning instruments can provide very high classification accuracy. In particular, support vector machines with proper optimization and developed local feature set gives competitive results. Secondly, a novel architecture of convolutional neural network is proposed. Important practical aspects of both methods are analyzed.
Development of indoor navigation services and algorithms is becoming a popular trend in IT-industry in recent years. Some of the modern buildings, like airports, shopping malls, warehouses have grown enough (Fig.1) to feel a need for their own navigation tools for customers. Closed environment conditions exclude the usage of common satellite-based navigation systems like GPS or GLONASS, so nowadays some alternative information sources of user localization appear at the scene. Fig.1.