Temporal Features Research Articles

Abnormal nonlinear features of EEG microstate sequence in obsessive–compulsive disorder

BackgroundAt present, only a few studies have explored electroencephalography (EEG) microstates of patients with obsessive–compulsive disorder (OCD) and the results are inconsistent. Additionally, the nonlinear features of EEG microstate sequences contain rich information about the brain, yet how the nonlinear features of EEG microstate sequences abnormally change in patients with OCD is still unknown.MethodsResting-state EEG data were collected from 48 OCD patients and macheted 48 healthy controls (HC). Subsequently, EEG microstate analysis was used to extract the microstate temporal parameters (duration, occurrence, coverage) and nonlinear features of EEG microstate sequences (sample entropy, Lempel–Ziv complexity, Hurst index). Finally, the temporal parameters and nonlinear features of EEG microstate sequences were sent to three kinds of machine learning models to classify OCD patients.ResultsBoth groups obtained four typical EEG microstate topographies. The duration of microstates A, B, and C in OCD patients decreased significantly, while the occurrence of microstate D increased significantly compared to HC. Sample entropy and Lempel–Ziv complexity of microstate sequences in OCD patients increased significantly, while Hurst index decreased significantly compared to HC. The classification accuracy using the nonlinear features of microstate sequences reached up to 85%, significantly higher than that based on microstate temporal parameter models.ConclusionThis study provides supplementary findings on EEG microstates in OCD patients with a larger sample size. We found that the nonlinear features of EEG microstate sequences in OCD patients can serve as potential electrophysiological biomarkers for distinguishing OCD patients.

An adaptive assessment method of power system transient stability considering PMU data loss

AbstractTransient stability assessment (TSA) plays an important role in ensuring the reliable operation of power systems. With the popularity of phasor measurement units (PMUs), data‐driven TSA methods have been widely concerned. However, the performance of TSA model may deteriorate when data loss occurs due to PMU failure. This paper proposes an adaptive assessment method for transient stability of power systems considering PMU data loss. First, considering the importance of temporal features, a collection of PMU clusters is constructed to minimize the failure risk and maintain full observability of the whole buses of the grid. Secondly, a weighted integrated assessment model based on PMU clusters is constructed by using an improved eXplainable Convolutional neural network for Multivariate time series classification (XCM) as a TSA classifier. The model can make full use of time series information to carry out adaptive TSA and maintain the robustness of the assessment performance even when PMU failure occurs. Finally, it is verified in a modified IEEE 39‐bus system with wind power and solar power. The effect of the proposed method shows high accuracy and strong anti‐noise interference ability in case of data loss.

Enhancing video anomaly detection for human suspicious behavior through deep hybrid temporal spatial network

Abnormal behavior exhibited by individuals with particular intentions is common, and when such behavior occurs in public places, it can cause physical and mental harm to others. Considering the rise in the automated approach for anomaly detection in videos, accuracy becomes essential. Most existing models follow a deep learning architecture, which faces challenges due to variations in motion. This research work develops a deep learning based hybrid architecture with temporal and spatial features. The hybrid temporal spatial network (HTSNet) consists of two customized architectures: a graph neural network (GNN) and a convolutional neural network (CNN). HTSNet combined with a novel classifier to extract features and classify normal and abnormal behavior. The performance of HTSNet is rigorously evaluated using the University of California, San Diego-Pedestrian 1 (UCSD Ped1) dataset, a benchmark in computer vision research for anomaly detection in video surveillance. The effectiveness of HTSNet is demonstrated through a comparative analysis with current state-of-the-art methods, using the area under the curve (AUC) metric as a standard measure of performance. This paper contributes to the advancement of video surveillance technology, providing a robust framework for enhancing public safety and security in an increasingly digital world.

Deep feature synthesis approach using selective graph attention for replay attack voice spoofing detection

<span lang="EN-US">As voice-based authentication becomes increasingly integrated into security frameworks, establishing effective defenses against voice spoofing, particularly replay attacks, is more crucial than ever. This paper presents a novel comprehensive framework for replay attack detection that leverages the integration of advanced spectral-temporal feature extraction and graph-based feature processing mechanisms. The proposed system presents the design of a waveform encoder and a novel temporal residual unit for spectral and temporal feature extraction in synchronous. Further, an approach of selective attention graph followed by multi-scale feature synthesis is employed to retain precise and spoof indicative feature vectors at the classification layer. The proposed method addresses the significant challenge of distinguishing genuine speech from replayed recordings. The validation of the proposed model is done on the ASVSpoof2019 dataset to demonstrate the efficacy of the proposed approach. The proposed system outperforms existing methods, achieving a lower equal error rate (EER) of 0.015 and a reduced tandem detection cost function (t-DCF) of 0.503. The comparative outcome exhibits the robustness of the method in identifying replay attacks.</span>

AttentionEP: Predicting Essential Proteins via Fusion of Multiscale Features by Attention Mechanisms

Identifying essential proteins is of utmost importance in the field of biomedical research due to their essential functions in cellular activities and their involvement in mechanisms related to diseases. In this research, a novel approach called AttentionEP for predicting essential proteins (EP) is introduced by attention mechanisms. This method leverages both cross-attention and self-attention frameworks, focusing on enhancing prediction accuracy through the integration of features across diverse scales. Spatial characteristics of proteins are obtained from the protein-protein interaction (PPI) network by employing Graph Convolutional Networks (GCN) and Graph Attention Networks (GAT). Following this, Bidirectional Long Short-Term Memory networks (BiLSTM) are employed to derive temporal features from gene expression datasets. Furthermore, spatial characteristics are derived by integrating data on subcellular localization with the application of Deep Neural Networks (DNN). In order to effectively integrate features across multiple scales, initial steps involve the application of self-attention techniques to derive essential insights from each unique data set. Following this, mechanisms involving self-attention and cross-attention are employed to enhance the interaction between diverse information sources. To identify essential proteins, a classifier based on the ResNet architecture is developed. The findings from the experiments indicate that the method introduced here shows superior performance in identifying essential proteins, recording an Area Under the Curve (AUC) value of 0.9433. This approach shows a considerable advantage over established techniques. The findings of this study provide a significant advancement in the comprehension of critical proteins, revealing promising potential for applications in the development of therapeutics and addressing various diseases.

A Data-Driven Approach for Stylolite Detection

Summary Stylolite is a specific geopattern that can occur in both sedimentary rocks and deformed zones, which could change the porosity of the reservoir, modify the permeability, and even result in horizontal permeability barriers. Though there are many related studies to characterize this geopattern, most of them focus on experimental methods. In this work, we investigated a new approach for recovering geometrical information of the stylolite zone (including its size and location) based on neural network architectures including convolutional neural network (CNN), long short-term memory (LSTM), and transformer encoder, which could serve as a data-driven solution to the problem. To our knowledge, this paper is the first to exclusively use well testing data for deducing field geopatterns, whereas other studies have relied on additional data sources such as well logging data. To simplify the problem, we first conducted simulation by building 3D multilayer reservoir models with one stylolite zone. We considered both simplified cases with only a few homogeneous layers and cases with heterogeneous layers to generalize our work. For the heterogeneous case, we extracted the permeability from SPE10 Model 2, a commonly used public resource (SPE 10 Benchmark, Model 2 2008). Producing and observing wells in the model are at different locations and provide pressure and production rate data as inputs for the deep learning models, in the form of multivariant time series data. For homogeneous cases, after zero-padding and standardizing our inputs to tackle different-length data and features with different scales, we applied a CNN-LSTM model to our data set, leveraging the CNN’s ability to capture short-term local features and the LSTM’s capacity to extract long-term dependencies. This combination improves the extraction of important information from time series data related to stylolites. The two subnetworks are connected in parallel to combine the advantages of CNN in extracting local temporal features and the strengths of LSTM in capturing long-time dependency via self-loops. Our work also separately covers the two subnetworks of the CNN-LSTM model as baseline models. For heterogeneous cases, a CNN-based model U-Net and an attention-based model set function for time series (SeFT) were introduced to make the predictions. In a more realistic scenario featuring fluid pathways with irregular shapes within a heterogeneous reservoir, we employed a transformer encoder model to predict both the shape and location of the fluid pathways. On the homogeneous data set, our CNN-LSTM model achieved a satisfactory performance and could predict the locations and sizes of the stylolite zone and outperformed the two baseline models. On the more challenging heterogeneous data set, our baseline and CNN-LSTM models failed to deliver meaningful results. In contrast, SeFT and U-Net performed better in the sense that we could successfully predict the locations of the stylolite zones. In a more realistic scenario involving irregularly-shaped fluid pathways, our transformer encoder model achieved accurate predictions of their locations.

Patient activity recognition for robust performance under varied illumination and cluttered background in indoor environment

Human Activity Recognition (HAR) holds significant importance in health and human-machine interaction. However, recognizing actions from 2D information faces challenges like occlusion, illumination variation, cluttered backgrounds, and view invariance. These hurdles are particularly pronounced in indoor patient monitoring settings due to fluctuating lighting conditions and cluttered backgrounds, which compromise the accuracy of activity recognition systems. A new architecture named illuminationRevive has been proposed to tackle this issue. Which utilizes an encoder-decoder convolutional neural network and image post-processing blocks to enhance the image's visual appearance. A new dataset comprising seven indoor physical activities has been proposed, created with contributions from thirty individuals aged 20 to 45. A hybrid fusion architecture is proposed to classify activities, integrating motion sequence information and body joint features. The proposed classification model incorporates generated Skeleton Motion History Images (SMHIs), collected human joint motion features from video frames, and novel kinematic and geometric features within window frames as inputs. The model can extract spatial and temporal feature vectors by integrating ResNet50-ViT (Residual Network-50 layers, Vision Transformer) and CNN-BiLSTM (Convolutional Neural Network-Bidirectional Long Short-Term Memory) layers. The suggested classification model was evaluated alongside state-of-the-art models using the LNMIIT-SMD (The LNM Institute of Information Technology-Skelton Motion Dataset) and established NTU-RGBD (Nanyang Technological University's Red Blue Green and Depth information) dataset. The evaluation aimed to assess the effectiveness of the proposed classification model architecture. Results demonstrate the superiority of the proposed model, achieving impressive accuracies of 98.21% on real-time data, 98.45% on the proposed dataset, and 97.12% on the NTU-RGBD dataset.

Enhancing Video-Language Representations With Structural Spatio-Temporal Alignment.

While pre-training large-scale video-language models (VLMs) has shown remarkable potential for various downstream video-language tasks, existing VLMs can still suffer from certain commonly seen limitations, e.g., coarse-grained cross-modal aligning, under-modeling of temporal dynamics, detached video-language view. In this work, we target enhancing VLMs with a fine-grained structural spatio-temporal alignment learning method (namely Finsta). First of all, we represent the input texts and videos with fine-grained scene graph (SG) structures, both of which are further unified into a holistic SG (HSG) for bridging two modalities. Then, an SG-based framework is built, where the textual SG (TSG) is encoded with a graph Transformer, while the video dynamic SG (DSG) and the HSG are modeled with a novel recurrent graph Transformer for spatial and temporal feature propagation. A spatial-temporal Gaussian differential graph Transformer is further devised to strengthen the sense of the changes in objects across spatial and temporal dimensions. Next, based on the fine-grained structural features of TSG and DSG, we perform object-centered spatial alignment and predicate-centered temporal alignment respectively, enhancing the video-language grounding in both the spatiality and temporality. We design our method as a plug&play system, which can be integrated into existing well-trained VLMs for further representation augmentation, without training from scratch or relying on SG annotations in downstream applications. On 6 representative VL modeling tasks over 12 datasets in both standard and long-form video scenarios, Finsta consistently improves the existing 13 strong-performing VLMs persistently, and refreshes the current state-of-the-art end task performance significantly in both the fine-tuning and zero-shot settings.

AutoLDT: a lightweight spatio-temporal decoupling transformer framework with AutoML method for time series classification

Time series classification finds widespread applications in civil, industrial, and military fields, while the classification performance of time series models has been improving with the recent development of deep learning. However, the issues of feature extraction effectiveness, model complexity, and model design uncertainty constrain the further development of time series classification. To address the above issues, we propose a Lightweight Spatio-Temporal Decoupling Transformer framework based on Automated Machine Learning technique (AutoLDT). The framework proposes a novel lightweight Transformer with fuzzy position encoding, TS-separable linear self-attention mechanism, and convolutional feedforward network, which mine the temporal and spatial features, as well as the local and global relationship of time series. Fuzzy positional encoding integrates fuzzy ideas to enhance the generalization performance of model information mining. TS-separable linear self-attention mechanism and convolutional feedforward network achieve feature extraction in a lightweight way by decoupling temporal and spatial features of time series. Notably, we adopt the Covariance Matrix Adaptation Evolution Strategy and global adaptive pruning technique to realize automated network structure design, which further improves the model training efficiency and automation, and avoids the uncertainty problem of network design. Finally, we validate the effectiveness of the proposed framework on publicly available UCR and UEA time series datasets. The experimental results show that the proposed framework not only improves the model classification performance in a lightweight way but also dramatically improves the model training efficiency.

VNN-DM: a vector neural network-based detection model for time synchronization attacks in park-level energy internet

Micro phasor measurement units (µPMUs) provide high-precision voltage and current phasor data, allowing real-time state estimation and fault detection, which are critical for the stability and reliability of modern power systems. However, their reliance on accurate time synchronization makes them vulnerable to time synchronization attacks (TSAs), which can disrupt grid monitoring and control by corrupting µPMU data. Addressing these vulnerabilities is essential to ensure the secure and resilient operation of smart grids and energy internet technologies. To address these challenges, intelligent detection methods are essential. Therefore, this paper proposes a µPMU measurement data TSA detection model based on vector neural networks (VNNs). This model initially employs a vector neural network to process raw data, effectively extracting and analyzing temporal features. During the same time, a capsule network is employed to classify these temporal features. On this basis, a reconstruction network is used to verify the representational capacity of the model. Simulations based on µPMU measurement data demonstrate that the model exhibits excellent detection capacity in various performance metrics, underscoring its precision and robustness.

An enhanced interval-valued PM2.5 concentration forecasting model with attention-based feature extraction and self-adaptive combination technology

Accurate predictions of air quality enable governments and relevant authorities to take promptly measures for protecting public health. With the increasing time-varying nature of air pollutants, predicting daily average concentrations alone is no longer sufficient for environmental management and risk warning. Hence, this paper proposes a multi-resolution interval-valued PM2.5 concentration combination prediction model, which based on interval decomposition and attention mechanism reconstruction. Firstly, the interval-valued time series (ITS) was decomposed and adaptively reconstructed using the binary empirical mode decomposition (BEMD) algorithm and attention-based reconstruction. Subsequently, multi-resolution linear projection layers were applied to extract temporal features from the time series. Finally, a hybrid prediction module was implemented that combines CNN and LSTM to predict each subsequence and integrate them to derive the final interval prediction values for PM2.5. In the proposed framework, the reconstruction technique effectively resolved the issue of inconsistent numbers of different feature decomposition subsequences, while the linear projection layer fully captured the multi-resolution characteristics of the time series. Empirical studies conducted in three districts of Beijing showed that, compared to state-of-the-art baseline models, the framework reduced the average values of five interval evaluation metrics by 11.2%, 17.4%, 11.7%, 10.5%, and 14.8%, respectively. This interval-valued prediction framework can effectively assist urban air quality management and warning.

A Multi-Source Data-Driven Analysis of Building Functional Classification and Its Relationship with Population Distribution

Buildings, as key factors influencing population distribution, have various functional attributes. Existing research mainly focuses on the relationship between land functions and population distribution at the macro scale, while neglecting the finer-grained, micro-scale impact of building functionality on population distribution. To address this issue, this study integrates multi-source geospatial and spatio-temporal big data and employs the XGBoost algorithm to classify buildings into five functional categories: residential, commercial, industrial, public service, and landscape. The proposed model innovatively incorporates texture, geometric, and temporal features of building images, as well as socio-economic characteristics extracted using the distance decay algorithm. The results yield the following conclusions: (1) The proposed method achieves an overall classification accuracy of 0.77, which is 0.12 higher than that of the random forest-based approach. (2) The introduction of time features and the distance decay method further improved the model performance, increasing the accuracy by 0.04 and 0.03, respectively. (3) The correlation between the building functions and population distribution varies significantly across different scales. At the district and county levels, residential, commercial, and industrial buildings show a strong correlation with population distribution, whereas this correlation is relatively weak at the street scale. This study advances the understanding of building functions and their role in shaping population distribution, providing a robust framework for urban planning and population modeling.

MACNet: A Multidimensional Attention-Based Convolutional Neural Network for Lower-Limb Motor Imagery Classification

Decoding lower-limb motor imagery (MI) is highly important in brain–computer interfaces (BCIs) and rehabilitation engineering. However, it is challenging to classify lower-limb MI from electroencephalogram (EEG) signals, because lower-limb motions (LLMs) including MI are excessively close to physiological representations in the human brain and generate low-quality EEG signals. To address this challenge, this paper proposes a multidimensional attention-based convolutional neural network (CNN), termed MACNet, which is specifically designed for lower-limb MI classification. MACNet integrates a temporal refining module and an attention-enhanced convolutional module by leveraging the local and global feature representation abilities of CNNs and attention mechanisms. The temporal refining module adaptively investigates critical information from each electrode channel to refine EEG signals along the temporal dimension. The attention-enhanced convolutional module extracts temporal and spatial features while refining the feature maps across the channel and spatial dimensions. Owing to the scarcity of public datasets available for lower-limb MI, a specified lower-limb MI dataset involving four routine LLMs is built, consisting of 10 subjects over 20 sessions. Comparison experiments and ablation studies are conducted on this dataset and a public BCI Competition IV 2a EEG dataset. The experimental results show that MACNet achieves state-of-the-art performance and outperforms alternative models for the subject-specific mode. Visualization analysis reveals the excellent feature learning capabilities of MACNet and the potential relationship between lower-limb MI and brain activity. The effectiveness and generalizability of MACNet are verified.

Using LSTM with Trajectory Point Correlation and Temporal Pattern Attention for Ship Trajectory Prediction

Accurate ship trajectory prediction is crucial for real-time vessel position tracking and maritime safety management. However, existing methods for ship trajectory prediction encounter significant challenges. They struggle to effectively extract long-term and complex spatial–temporal features hidden within the data. Moreover, they often overlook correlations among multivariate dynamic features such as longitude (LON), latitude (LAT), speed over ground (SOG), and course over ground (COG), which are essential for precise trajectory forecasting. To address these pressing issues and fulfill the need for more accurate and comprehensive ship trajectory prediction, we propose a novel and integrated approach. Firstly, a Trajectory Point Correlation Attention (TPCA) mechanism is devised to establish spatial connections between trajectory points, thereby uncovering the local trends of trajectory point changes. Subsequently, a Temporal Pattern Attention (TPA) mechanism is introduced to handle the associations between multiple variables across different time steps and capture the dynamic feature correlations among trajectory attributes. Finally, a Great Circle Route Loss Function (GCRLoss) is constructed, leveraging the perception of the Earth’s curvature to deepen the understanding of spatial relationships and geographic information. Experimental results demonstrate that our proposed method outperforms existing ship trajectory prediction techniques, showing enhanced reliability in multi-step predictions.

MGSAN: multimodal graph self-attention network for skeleton-based action recognition

Due to the emergence of graph convolutional networks (GCNs), the skeleton-based action recognition has achieved remarkable results. However, the current models for skeleton-based action analysis treat skeleton sequences as a series of graphs, aggregating features of the entire sequence by alternately extracting spatial and temporal features, i.e., using a 2D (spatial features) plus 1D (temporal features) approach for feature extraction. This undoubtedly overlooks the complex spatiotemporal fusion relationships between joints during motion, making it challenging for models to capture the connections between different temporal frames and joints. In this paper, we propose a Multimodal Graph Self-Attention Network (MGSAN), which combines GCNs with self-attention to model the spatiotemporal relationships between skeleton sequences. Firstly, we design graph self-attention (GSA) blocks to capture the intrinsic topology and long-term temporal dependencies between joints. Secondly, we propose a multi-scale spatio-temporal convolutional network for channel-wise topology modeling (CW-TCN) to model short-term smooth temporal information of joint movements. Finally, we propose a multimodal fusion strategy to fuse joint, joint movement, and bone flow, providing the model with a richer set of multimodal features to make better predictions. The proposed MGSAN achieves state-of-the-art performance on three large-scale skeleton-based action recognition datasets, with accuracy of 93.1% on NTU RGB+D 60 cross-subject benchmark, 90.3% on NTU RGB+D 120 cross-subject benchmark, and 97.0% on the NW-UCLA dataset. Code is available at https://github.com/lizaowo/MGSAN.

Fast POI anomaly detection using a weakly-supervised temporal state regression network

Point-of-interest (POI) is a fundamental data type of maps. Anomalous POIs would make maps outdated and lead to user-unfriendly location-based services, and thus should be discovered as fast as possible. Traditional POI anomaly detection methods are inefficient owing to high investigation costs. The emergence of massive human activity data provides a new insight into monitoring POI states through time series modeling. When a POI turns into an anomaly, the associated human activity would disappear. However, human activity data have complicated temporal patterns and noises. It is challenging for existing time series methods to model human activity dynamics. More importantly, there is a lag between the time a POI becomes anomalous and the time we discover it. In this research, we develop a temporal state regression network (TSRNet) model for fast POI anomaly detection. The model can extract temporal features in human activity data, and predict POI state scores as anomaly indicators. Meanwhile, an inference approach is proposed to generate state score sequences as inexact labels for model training. Such weak labels enable TSRNet to identify abnormal temporal patterns as soon as they appear, so that POI outliers can be detected at an early time. Experiments on real-word datasets from AMAP validate the feasibility of our method.

An automatic approach to detect skin cancer utilizing active infrared thermography

Skin cancer is a growing global concern, with cases steadily rising. Typically, malignant moles are identified through visual inspection, using dermatoscopy and patient history. Active thermography has emerged as an effective method to distinguish between malignant and benign lesions. Our previous research showed that spatio-temporal features can be extracted from suspicious lesions to accurately determine malignancy, which was applied in a distance-based classifier.In this study, we build on that foundation by introducing a set of novel spatial and temporal features that enhance classification accuracy and can be integrated into any machine learning approach. These features were implemented in a support-vector machine classifier to detect malignancy. Notably, our method addresses a common limitation in existing approaches—manual lesion selection—by automating the process using a U-Net convolutional neural network.We validated our system by comparing U-Net's performance with expert dermatologist segmentations, achieving a 17% improvement in the Jaccard index over a semi-automatic algorithm. The detection algorithm relies on accurate lesion segmentation, and its performance was evaluated across four segmentation techniques. At an 85% sensitivity threshold, expert segmentation provided the highest specificity at 87.62%, while non-expert and U-Net segmentations achieved comparable results of 69.63% and 68.80%, respectively. Semi-automatic segmentation lagged behind at 64.45%. This automated detection system performs comparably to high-accuracy methods while offering a more standardized and efficient solution. The proposed automatic system achieves 3% higher accuracy compared to the ResNet152V2 network when processing low-quality images obtained in a clinical setting.

A novel model for mapping soil organic matter: Integrating temporal and spatial characteristics

Mapping the spatial distribution of soil organic matter (SOM) content is crucial for land management decisions, yet its accurate mapping faces challenges due to complex soil-environment relationships and temporal feature capture limitations in machine learning models. This study focuses on the typical black soil region in Northeast China, specifically using Youyi Farm as the main research area and Heshan Farm as the transfer research area. A novel approach is proposed that combines the CNN-LSTM model with a Cosine Annealing Warm Restarts learning rate (CNN-LSTM-CAWR) to enhance the accuracy of SOM mapping. In this model, the Convolutional Neural Network (CNN) extracts spatial context features from static variables (e.g., climate and terrain variables), while the Long Short-Term Memory (LSTM) network captures temporal features from dynamic variables (e.g., Sentinel-2 time series from April to October). The incorporation of the CAWR learning rate helps alleviate overfitting issues. Comparing the CNN-LSTM model, CNN model, and traditional RF model, the results show that the CNN-LSTM-CAWR model achieves the highest accuracy within research Area 1 (R2 = 0.64, RMSE = 0.54 %) and maintains strong performance in the transfer research area (R2 = 0.60, RMSE = 0.57 %). CNN-LSTM-CAWR demonstrates faster convergence, thereby improving mapping precision and effectively utilizing temporal information from features to enhance overall model performance. This study underscores the significant potential of the hybrid CNN-LSTM with CAWR model, highlighting the valuable information for SOM mapping contained within Sentinel-2 time series data.

A Review on Computer Vision-Based Methods for Abnormal Human Action Recognition

Human behavior recognition constitutes a crucial research domain within both computer vision and behavior recognition. As a branch of human behavior recognition, abnormal behavior recognition has witnessed rapid advancement in recent years, which is capable of enhancing the governance level of public safety. To investigate the theoretical and technical progress of abnormal behavior recognition in public places within the realm of computer vision, this paper initially delineates the definition of abnormal behavior in public places. Secondly, as computer vision and pattern recognition technologies have progressed, algorithms are now divided into two distinct categories: traditional methods and those leveraging deep learning techniques. The identification of atypical human behavior can be divided into two approaches depending on the detection of body key points: one approach uses skeletal key points, while the other focuses on analyzing temporal and spatial features. Finally, this paper conducts a review of the mainstream datasets of abnormal human behavior both at home and abroad, analyzes the performance of related algorithms on the datasets, and gives future research directions as well as optimization suggestions.

Improved LSTM-Squeeze Net Architecture for brain activity detection using EEG with improved feature set

Brain activity recognition (BAR) has emerged as most intriguing fields of research in recent days. It poses the ability to transform an extensive number of appliances, together with ICU surveillance, appliance control for the disabled and aged, and the detection of neural diseases. EEG data are the most common representation of brain activity, as they capture the voltage variations of neurons in the brain using non-invasive electrodes implanted on the scalp. Deep learning models play a major role in recognizing brain activity. This work uses EEG inputs to present a new hybrid model-based BAR. Adaptive least mean square (ALMS) is first used to filter the input signal for EEG signals. Next, improved CSP features, morphological features, temporal domain features, and wavelet packet decomposition (WPD) features are extracted. Appropriate characteristics are chosen following feature extraction using Improved Mutual Information (IMI). Then, using an enhanced LSTM (ILSTM)-Squeeze net model, brain actions are detected or recognized. The highest accuracy at LP = 90 is around 0.961. In contrast, the precisions of 0.7385, 0.667, 0.7516, 0.771, 0.686, 0.7124, and 0.7189 were obtained by Dense Net, Mobile Net, Squeeze Net, Res Net, ANN, LSTM, and RNN. The evaluation done proves the efficiency of the proposed model over the other conventional models.