A lightweight false alarm suppression method in heterogeneous change detection

Overlooking the issue of false alarm suppression in heterogeneous change detection leads to inferior detection per-formance.This paper proposes a method to handle false alarms in heterogeneous change detection.A lightweight network of two channels is bulit based on the combination of convolutional neural network(CNN)and graph convolutional network(GCN).CNNs learn feature difference maps of multitemporal images,and attention modules adaptively fuse CNN-based and graph-based features for different scales.GCNs with a new kernel filter adaptively distinguish between nodes with the same and those with different labels,generating change maps.Experimental evaluation on two datasets validates the efficacy of the pro-posed method in addressing false alarms.

Bayesian track-before-detect algorithm for nonstationary sea clutter

Radar detection of small targets in sea clutter is a particularly demanding task because of the nonstationary characteristic of sea clutter.The track-before-detect(TBD)filter is an effective way to increase the signal-to-clutter ratio(SCR),thus improving the detection performance of small targets in sea clutter.To cope with the nonstationary characteristic of sea clutter,an easily-implemented Bayesian TBD filter with adaptive detection threshold is proposed and a new parameter estimation method is devised which is integrated into the detection process.The detection threshold is set according to the parameter estimation result under the framework of information theory.For detection of closely spaced targets,those within the same range cell as the one under test are treated as contribution to sea clutter,and a successive elimination method is adopted to detect them.Simulation results prove the effectiveness of the proposed algorithm in detecting small targets in nonstationary sea clutter,especially closely spaced ones.

Research on MLChecker Plagiarism Detection System

Plagiarism detection system plays an essential role in education quality improvement by helping teachers to detect plagiarism.Using a number of measures customized to determine occurrences of plagiarism is the most common approach for plagiarism detection tool.It is simple and effective,while it lacks flexibility when applied in more complicated situations.This paper proposes the MLChecker,a smart plagiarism detection system,to provide more flexible detection tactics.An automatic plagiarism dataset construction method was exploited in MLChecker to dynamically update the plagiarism detection algorithms according to different detection tasks.And the full-process quality management functions were also provided by MLChecker.The result shows that the detection accuracy is raised by 56%.Compared with traditional plagiarism detection tools,MLChecker is with higher accuracy and efficiency.