With the continuous development of the convolutional neural network (CNN) concept and other deep learning technologies, target recognition in Synthetic Aperture Radar (SAR) images has entered a new stage. At present, shallow CNNs with simple structure are mostly applied in SAR image target recognition, even though their feature extraction ability is limited to a large extent. What’s more, research on improving SAR image target recognition efficiency and imbalanced data processing is relatively scarce. Thus, a lightweight CNN model for target recognition in SAR image is designed in this paper. First, based on visual attention mechanism, the channel attention by-pass and spatial attention by-pass are introduced to the network to enhance the feature extraction ability. Then, the depthwise separable convolution is used to replace the standard convolution to reduce the computation cost and heighten the recognition efficiency. Finally, a new weighted distance measure loss function is introduced to weaken the adverse effect of data imbalance on the recognition accuracy of minority class. A series of recognition experiments based on two open data sets of MSTAR and OpenSARShip are implemented. Experimental results show that compared with four advanced networks recently proposed, our network can greatly diminish the model size and iteration time while guaranteeing the recognition accuracy, and it can effectively alleviate the adverse effects of data imbalance on recognition results.

No abstract

The genetic algorithm (GA) often suffers from the premature convergence because of the loss of population diversity at an early stage of searching. This paper proposes a steep thermodynamical evolutionary algorithm (STEA), which utilizes a steep thermodynamical selection (STS) rule. STEA simulates the competitive mechanism between energy and entropy in annealing to systematically resolve the conflicts between selective pressure and population diversity in GA. This paper also proves that the rule STS has the approximate steepest descent ability of the free energy. Experimental results show that STEA is both far more efficient and much stabler than the thermodynamical genetic algorithm (TDGA).

With the development of e-learning, many authoring tools have been designed and implemented, besides of course authoring tools like eXe, presentation tools like PowerPoint are widely used and play important role in creating learning content. However, most presentation tools do not support PowerPoint files which are very popular in the education because of its complexity. This will lead to poor shareability of existed PowerPoint files. As a result, PowerPoint file converting should be taken into account for achieving interoperability among presentation tools. This paper describes the initial efforts to support PowerPoint converting of a similar presentation tool named MCAT which has been implemented. The file format of MCAT is based of xml. It makes MCAT support PowerPoint files by converting PowerPoint file to xml file. Further, the converted PowerPoint file can be used in three-screen streaming courseware to reduce needed bandwidth by transforming xml file to html file using playing tool.

No abstract