Electroencephalogram (EEG) plays an important role in identifying brain activity and behavior. However, the recorded electrical activity always be contaminated with artifacts and then affect the analysis of EEG signal. Hence, it is essential to develop methods to effectively detect and extract the clean EEG data during encephalogram recordings. Several methods have been proposed to remove artifacts, but the research on artifact removal continues to be an open problem. This paper tends to review the current artifact removal of various contaminations. We first discuss the characteristics of EEG data and the types of different artifacts. Then, a general overview of the state-of-the-art methods and their detail analysis are presented. Lastly, a comparative analysis is provided for choosing a suitable methods according to particular application.

The Internet of Things (IoT) oriented intelligent services require high-quality sensor data delivery in the wireless sensor networks (WSNs). However, either due to the sensor malfunctions and commutation errors or simply due to the expensive overhead for making full data forwarding, data corruption and loss is relatively common in WSNs, which adversely affects the data quality and the further decisions taking from data. Motivated by the emerging field of graph signal processing (GSP), we propose to impute the missing values in wireless sensor networks based on the topological information carried in the product graph. The proposed solution captures the joint time-space dependencies among the sensor data through a spatiotemporal (ST) graph, which is a time-vertex graph constructed by taking a strong product of a temporal graph and a spatial sensor network graph. Then, the sensory data are mapped onto the vertices of the ST graph and the spatial-temporal nature of sensor data can be further characterized by the notion of smoothness used in GSP. Moreover, instead of imputing with a given spatial graph, we propose a graph learning-based imputation framework to infer underlying space dependencies between the sensors and thus enhance the data imputation performances. Finally, we validate the proposed recovery method using real-world sensor network datasets. The results demonstrate the superior performance of our proposed graph-based method in sensor data imputation, especially when massive sensor data are lost.

Over the past few decades, plenty of visualization software for the structural analysis of disordered/complex systems has been developed, but the uniqueness and correctness of structural quantification for such systems are still challenging. This paper introduces a visualization analysis tool based on the largest standard cluster analysis (LaSCA), which satisfies the three essential requirements for general structural analysis: physical correctness, objective identification, and injective representation. The specific functionalities of LaSCA include the directed graph model of complex systems, novel structural parameters, topologically close-packed structures, arbitrary partial pair distribution functions, the identification of long-range ordered structures, the adaptive selection of graphical elements, the tracking display of atom ID, user-defined view angles, various options for atom selection, and so on. The program is efficiently based on OpenGL hardware acceleration, employing special algorithms to treat bonds as cylinders or lines and treat atoms as spheres, icosahedrons, tetrahedrons, or points. LaSCA can process more than 1.2 million atoms within 50 s on a PC with 1 GB memory and four cores (Intel Core i7-9700). It is robust and low-cost for surveying short-, medium-, and long-range ordered structures and tracking their evolutions.

Modeling the dynamics of glassy systems has been challenging in physics for several decades. Recent studies have shown the efficacy of Graph Neural Networks (GNNs) in capturing particle dynamics from the graph structure of glassy systems. However, current GNN methods do not take the dynamic patterns established by neighboring particles explicitly into account. In contrast to these approaches, this paper introduces a novel dynamical parameter termed “smoothness” based on the theory of graph signal processing, which explores the dynamic patterns from a graph perspective. Present graph-based approaches encode structural features without considering smoothness constraints, leading to a weakened correlation between structure and dynamics, particularly on short timescales. To address this limitation, we propose a Geometry-enhanced Graph Neural Network (Geo-GNN) to learn the smoothness of dynamics. Results demonstrate that our method outperforms state-of-the-art baselines in predicting glassy dynamics. Ablation studies validate the effectiveness of each proposed component in capturing smoothness within dynamics. These findings contribute to a deeper understanding of the interplay between glassy dynamics and static structure.