High False Alarm Rate Research Articles

FireNet: A Lightweight and Efficient Multi-Scenario Fire Object Detector

Fire and smoke detection technologies face challenges in complex and dynamic environments. Traditional detectors are vulnerable to background noise, lighting changes, and similar objects (e.g., clouds, steam, dust), leading to high false alarm rates. Additionally, they struggle with detecting small objects, limiting their effectiveness in early fire warnings and rapid responses. As real-time monitoring demands grow, traditional methods often fall short in smart city and drone applications. To address these issues, we propose FireNet, integrating a simplified Vision Transformer (RepViT) to enhance global feature learning while reducing computational overhead. Dynamic snake convolution (DSConv) captures fine boundary details of flames and smoke, especially in complex curved edges. A lightweight decoupled detection head optimizes classification and localization, ideal for high inter-class similarity and small targets. FireNet outperforms YOLOv8 on the Fire Scene dataset (FSD) with a mAP@0.5 of 80.2%, recall of 78.4%, and precision of 82.6%, with an inference time of 26.7 ms. It also excels on the FSD dataset, addressing current fire detection challenges.

An intelligent monitoring approach for urban natural gas pipeline leak using semi-supervised learning generative adversarial networks

Traditional gas pipeline leak monitoring methods are subjected to the long response times and high false alarm rates. Deep learning can enhance the accuracy and real-time performance of pipeline leak monitoring. This paper develops an intelligent monitoring approach for urban gas pipeline leaks based on a semi-supervised learning Generative Adversarial Network (SGAN). First, the Isolation Forest algorithm is used to classify anomalies in the collected process parameter data of urban natural gas pipelines. One-Hot Encoding is used to label a small amount of sample data of pipeline leak. Second, both the labeled and unlabeled data are input into SGAN model for semi-supervised learning and classification to monitor the state of urban gas pipeline leak. The methodology addresses the imbalance between pipeline leak status data and normal data. The comparison with GAN and MLP shows that the methodology reaches the highest values in all evaluation metrics (precision = 94.1%, accuracy = 95.63%, recall = 93.93%, F1 score = 94.32%). The superior performance and accuracy make it more effective for urban natural gas pipeline leak monitoring.

Краткосрочный прогноз смерчеопасных ситуаций для прибрежной акватории Черного моря с использованием усовершенствованного индекса WRI

The paper considers the issues of improving the quality of short-term forecasts of waterspout-risk conditions for the Black Sea coastal water. The results related to the development of new versions for calculating the regional waterspout risk index (WRI) for both the warm (WRI21) and cold (WRIW) seasons are presented. The WRI is used to identify zones of high-risk waterspout formation and subsequent localization of waterspout-risk areas of the coast. Calculations of the WRI fields are based on the output of the COSMO-Ru2 mesoscale model with a grid spacing of 2.2 km. Calculations are carried out in real time within the technology for assessing the risk of waterspouts. The skill of forecasting new WRI versions for different periods is analyzed. It is shown that the probability of waterspout detection in the warm season can reach 82% on average. At the same time, there is also a high false alarm rate (about 70%), mainly due to the excessively predicted waterspout risk in certain areas of the coast (Sochi and Tuapse). In the cold season, as expected, a significantly smaller number of waterspout-risk days is predicted as compared to the warm period. Examples of forecasts for different seasons are presented. Recommendations on the use of forecasts for preparing storm warnings are given.

Condition assessment and predictive maintenance for contact probe using health index and encoder‐decoder LSTM model

AbstractContact probe is broadly used for the continuous monitoring of microelectronic components in manufacturing industries. False rejection of fine product due to defective contact probe significantly reduces the yield in production. Traditionally, defect detection for contact probes heavily depends on a valid range manually defined by engineers over the measured value of certain parameters. However, the subjective range defined according to engineer experience is prone to trigger a high rate of false alarms due to the inherent noise in the measured parameters. To address this issue, we construct a health index (HI) with the contact resistance‐directly‐related features to help monitor and assess the condition of contact probe. Based on the established HI, we develop Long Short‐Term Memory (LSTM) encoder‐decoder machine learning model to assess the condition of contact probe by forecasting the HI value in the future. Encoders from LSTM and convolutional neural network (CNN) are selected as the encoder‐decoder architecture for the sequence‐to‐sequence prediction due to their advantage in extracting the correlation of features at different scales. An explainable Artificial Intelligence (XAI) technique named Local interpretable model‐agnostic explanations (LIME) is used to quantify the contribution of each feature to the model prediction. The encoder from CNN is found to outperform the LSTM encoder in extracting the inter‐feature correlation. Finally, the predicted HI is used to signal the alarm for the maintenance action of contact probe when its value is below a predefined threshold. Comparison between the action alarm triggered by the developed HI and the actual maintenance records suggests that the proposed approach achieves at least 75% accuracy for the triggered alarm in the next 15 mins.

MFT: A novel memory flow transformer efficient intrusion detection method

Intrusion detection is a critical field in network security research that is devoted to detecting malicious traffic or attacks on networks. Even with the advances in today's Internet environment, a lot of intrusion detection techniques still fail to take into account the long-term characteristics present in network data, which results in a high false alarm rate. Some researchers have tried to address this problem by using the traditional transformer model; however, it is not very effective when dealing with complex relationships and the subtle classification requirements of large amounts of sequential data. This work presents a novel solution called the memory flow transformer (MFT) in response to the limitations of the conventional transformer model. By utilizing a carefully designed memory flow structure, MFT transcends traditional limitations and makes it possible to obtain complex long-term features from network traffic. This innovation enables the model to identify deep connections at a finer level between a wide variety of network traffic data. Extensive experiments were carried out on the complex CICIDS 2017 and NSL-KDD datasets to validate the effectiveness of the MFT model. The results were outstanding, demonstrating MFT's powerful detection abilities. With regard to performance metrics like accuracy, F1 score, false alarm rate, and training time, MFT is superior to current state-of-the-art approaches. Network security is greatly strengthened by MFT, which provides practitioners in the intrusion detection field with novel and effective techniques.

K-means adaptive 2DSSA based on sparse representation model for hyperspectral target detection

Target detection is a hot spot in hyperspectral imagery (HSI) processing. The detection accuracy of target detection algorithms based on sparse representation (SR) models usually suffers from the high reconstruction residuals caused by inaccurate background estimations and insufficient target samples. Besides, with the development of hyperspectral imaging technology, the spatial resolution of HSI has been continuously enhanced, which can provide more spatial information for target detection. However, spatial information is often overlooked, leading to the underutilization of the pluralistic features of HSI. Target detection using only spectral information is susceptible to spectral variation, resulting in a high false alarm rate. To alleviate these problems, this paper proposes a joint spatial-spectral algorithm. In terms of spectra, a dictionary construction strategy (DCS) is designed for the sparse representation-based binary hypothesis (SRBBH) detector to reduce reconstruction residuals of target and background samples. In terms of space, k-means 2D adaptive singular spectrum analysis (KSSA) is used to extract spatial features in cluster units. Using spatial features can enhance the robustness of the algorithm to spectral variation, thereby reducing false alarms. The target detection results are obtained by applying DCS-SRBBH to the KSSA feature image. We evaluate the proposed algorithm on three datasets: two public and one of our own. Comprehensive experimental results indicate that the proposed algorithm outperforms other target detection algorithms in terms of accuracy.

Advanced Hybrid Techniques for Cyberattack Detection and Defense in IoT Networks

ABSTRACTThe Internet of Things (IoT) represents a vast network of devices connected to the Internet, making it easier for users to connect to modern technology. However, the complexity of these networks and the large volume of data pose significant challenges in protecting them from persistent cyberattacks, such as distributed denial‐of‐service (DDoS) attacks and spoofing. It has become necessary to use intrusion detection systems and protect these networks. Existing intrusion detection systems for IoT networks face many problems and limitations, including high false alarm rates and delayed detection. Also, the datasets used for training may be outdated or sparse, which reduces the model's accuracy, and mechanisms may not be used to defend the network when any intrusion is detected. To address these limitations, a new hybrid deep learning and machine learning methodology is proposed that contributes to detecting DDoS and spoofing attacks, reducing false alarms, and then implementing the necessary defensive measures. In proposed hybrid methodology consists of three stages: the first stage is to propose a hybrid method for feature selection consisting of techniques (correlation coefficient and sequential feature selector); the second stage is to propose a hybrid model by integrating deep learning neural networks with a machine learning classifier (cascaded long short‐term memory [LSTM] and Naive Bayes classifier); and in the third stage, improving network defense mechanisms and blocking ports after detecting threats and maintaining network integrity. In training and evaluating the performance of the proposed methodology, three datasets (CIC‐DDoS2019, CIC‐IoT2023, and CIC‐IoV2024) were used, and these data were also balanced to obtain effective results. The accuracy of 99.91%, 99.88%, and 99.77% was obtained. Also, a cross‐validation technique was used with the test data to ensure no overfitting. The proposed methodology has proven its high performance in detecting attacks, as it provides a powerful solution to enhance the security of IoT networks and protect them from cyberattacks, as it can be applied in many fields and to other attacks.

A multiscale approach for network intrusion detection based on variance–covariance subspace distance and EQL v2

As an important network defense approach, network intrusion detection is mainly used to identify anomaly traffic behavior. However, dominant network intrusion detection approaches are now struggling to identify the complex and variable means of attack, leading to high false alarm rate. Additionally, the feature redundancy and class imbalance problem in the intrusion detection dataset also constrain the performance of detection methods. This paper proposes a multiscale intrusion detection approach based on variance–covariance subspace distance and Equalization Loss v2 (EQL v2). Firstly, the variance–covariance subspace distance is used for feature selection on the preprocessed dataset to determine a set of representative feature subsets that can effectively approximate the original feature space. Secondly, the loss function, EQL v2, is adopted to balance the positive and negative gradients, addressing the class imbalance problem. Finally, a pyramid depthwise separable convolution model is proposed to capture the multiscale information of the traffic, and the convolutional layer in the depthwise convolution is replaced with self-supervised predictive convolutional attention block to compensate for the performance loss caused by the parameter reduction. Extensive experiments demonstrated that the proposed approach exhibits better performance on the three datasets of NSL-KDD, UNSW_NB15, and CIC-IDS-2017, with accuracy rates of 99.19%, 97.81%, and 99.83%, respectively, effectively improve the intrusion detection performance.

Self-Supervised Machine Learning Framework for Online Container Security Attack Detection

Container security has received much research attention recently. Previous work has proposed to apply various machine learning techniques to detect security attacks in containerized applications. On one hand, supervised machine learning schemes require sufficient labeled training data to achieve good attack detection accuracy. On the other hand, unsupervised machine learning methods are more practical by avoiding training data labeling requirements, but they often suffer from high false alarm rates. In this article, we present a generic self-supervised hybrid learning (SHIL) framework for achieving efficient online security attack detection in containerized systems. SHIL can effectively combine both unsupervised and supervised learning algorithms but does not require any manual data labeling. We have implemented a prototype of SHIL and conducted experiments over 46 real-world security attacks in 29 commonly used server applications. Our experimental results show that SHIL can reduce false alarms by 33%–93% compared to existing supervised, unsupervised, or semi-supervised machine learning schemes while achieving a higher or similar detection rate.

Real-Time Fire Classification Models Based on Deep Learning for Building an Intelligent Multi-Sensor System

Fire detection systems are critical for mitigating the damage caused by fires, which can result in significant annual property losses and fatalities. This paper presents a deep learning-based fire classification model for an intelligent multi-sensor system aimed at early and reliable fire detection. The model processes data from multiple sensors that detect various parameters, such as temperature, humidity, and gas concentrations. Several deep learning architectures were evaluated, including LSTM, GRU, Bi-LSTM, LSTM-FCN, InceptionTime, and Transformer. The models were trained on data collected from controlled fire scenarios and validated for classification accuracy, loss, and real-time performance. The results indicated that the LSTM-based models (particularly Bi-LSTM and LSTM) could achieve high classification accuracy and low false alarm rates, demonstrating their effectiveness for real-time fire detection. The findings highlight the potential of advanced deep-learning models to enhance the reliability of sensor-based fire detection systems.

Ultra-lightweight convolution-transformer network for early fire smoke detection

BackgroundForests are invaluable resources, and fire is a natural process that is considered an integral part of the forest ecosystem. Although fire offers several ecological benefits, its frequent occurrence in different parts of the world has raised concerns in the recent past. Covering millions of hectares of forest land, these fire incidents have resulted in the loss of human lives, wild habitats, civil infrastructure, and severe damage to the environment. Around 90% of wildland fires have been caused by humans intentionally or unintentionally. Early detection of fire close to human settlements and wildlife centuries can help mitigate fire hazards. Numerous artificial intelligence-based solutions have been proposed in the past decade that prioritize the detection of fire smoke, as it can be caught through remote sensing and provide an early sign of wildland fire. However, most of these methods are either computationally intensive or suffer from a high false alarm rate. In this paper, a lightweight deep neural network model is proposed for fire smoke detection in images captured by satellites or other remote sensing sources.ResultsWith only 0.6 million parameters and 0.4 billion floating point operations per second, the hybrid network of convolutional and vision transformer blocks efficiently detects smoke in normal and foggy environmental conditions. It outperforms seven state-of-the-art methods on four datasets, including a self-collected dataset from the “Moderate Resolution Imaging Spectroradiometer” satellite imagery. The model achieves an accuracy of more than 99% on three datasets and 93.90% on the fourth dataset. The t-distributed stochastic neighbor embedding of extracted features by the proposed model demonstrates its superior feature learning capabilities. It is remarkable that even a tiny occurrence of smoke covering just 2% of the satellite image area is efficiently detected by the model.ConclusionsWith low memory and computational demands, the proposed model works exceedingly well, making it suitable for deployment in resource constrained devices for forest surveillance and early fire smoke detection.

End-to-End Learning-Based Study on the Mamba-ECANet Model for Data Security Intrusion Detection

With the rapid development of information technology, network security issues have become increasingly prominent. In particular, data security intrusions pose serious threats to the data privacy and system security of enterprises and individuals. Traditional intrusion detection systems often exhibit low detection accuracy and high false alarm rates when faced with complex and dynamic network environments and diverse attack methods. Therefore, this paper proposes a data security intrusion detection system based on deep learning, which integrates the Mamba model and ECANet model and employs an end-to-end learning approach for training and optimization. First, the Mamba model is introduced for preliminary data feature extraction, whose efficient feature representation capabilities provide a solid foundation for the subsequent detection process. Then, by integrating the ECANet model, feature selection is further optimized through the attention mechanism, enhancing the model’s focus on important features. Finally, an end-to-end learning approach is adopted to train and optimize the entire system, ensuring excellent performance and robustness in practical applications. Experimental results show that the proposed intrusion detection system demonstrates higher detection accuracy on multiple test datasets, improving by approximately 5% compared to traditional methods, providing a new and effective solution for data security.

Data-driven unsupervised anomaly detection of manufacturing processes with multi-scale prototype augmentation and multi-sensor data

Accurate anomaly detection (AD) of machine tools is crucial to ensure the quality and efficiency of the manufacturing processes. Due to the lack of tool anomaly information, it is difficult for AD model to precisely capture the distribution of health states and then obtain a discriminative decision boundary. Current methods try to reconstruct the normal data distribution without restricting the abnormal, resulting in the unacceptable overlap between normal and abnormal regions and finally leading to high false alarm rate. To tackle these issues, a hierarchical augmented autoencoder is proposed for AD of machine tools during manufacturing. First, a skip-connected autoencoder is built to basically learn the normal representations of multi-sensor data in an unsupervised manner. Then, to improve further emphasis the reconstruction on normality and suppress that on anomalies, we propose hierarchical memory modules to store multi-scale normal prototypical patterns, using them as a prior to guide the reconstruction with preference. Finally, A compound metric loss function is designed to measure data similarity considering both distance and angle perspectives, which can restrain noise interference and enhance model robustness. Extensive experiments are conducted on real-world CNC machine tool datasets, the proposed method achieves better performance for unsupervised AD compared with other typical methods.

High-Level Feature Fusion Deep Learning Model for Fault Detection in Handling Equipment in Dry Bulk Ports

The flexibility of handling equipment in dry bulk ports is poor, and frequent equipment fault induced by the high-load and high-power working conditions greatly impacts the overall port handling operations, making accurate fault detection play an important role in improving the efficiency and stability of dry bulk port operations. However, as we know, most fault detection methods for port handling equipment depend heavily on monitoring sensor data, which is not applicable in the dry bulk port due to high configuration and maintenance cost, as well as the high false alarm rate of monitoring sensors caused by strong background noise. To solve the problem, this study proposes a High-Level Feature Fusion Deep Learning Model, which uses different deep learning sub-models to extract features of structured and unstructured data. It fuses the extracted feature vectors to achieve fault detection in the handling equipment, establishing the mapping relationship between the fault (e.g., waiting for the pre-loading process, equipment fault, and others) and multi-source heterogeneous operation and maintenance data for the handling equipment, including reclaimers, belt conveyors, dumpers, and ship loaders. To verify the effectiveness of the proposed method, the actual data of a coal port in Northern China is employed as an example. The results show the deep learning model can achieve high prediction accuracy (over 86%) with high efficiency (0.5 s for each sample), which provides decision support for the fault detection in dry bulk port handling equipment.

Mutation-Based Multivariate Time-Series Anomaly Generation on Latent Space with an Attention-Based Variational Recurrent Neural Network for Robust Anomaly Detection in an Industrial Control System

Anomaly detection involves identifying data that deviates from normal patterns. Two primary strategies are used: one-class classification and binary classification. In Industrial Control Systems (ICS), where anomalies can cause significant damage, timely and accurate detection is essential, often requiring analysis of time-series data. One-class classification is commonly used but tends to have a high false alarm rate. To address this, binary classification is explored, which can better differentiate between normal and anomalous data, though it struggles with class imbalance in ICS datasets. This paper proposes a mutation-based technique for generating ICS time-series anomalies. The method maps ICS time-series data into a latent space using a variational recurrent autoencoder, applies mutation operations, and reconstructs the time-series, introducing plausible anomalies that reflect multivariate correlations. Evaluations of ICS datasets show that these synthetic anomalies are visually and statistically credible. Training a binary classifier on data augmented with these anomalies effectively mitigates the class imbalance problem.

Cross media knowledge information retrieval model based on D-S evidence theory

Cross media knowledge information retrieval provides strong support for information processing and utilization in the information society, but there are problems such as heterogeneity in cross media knowledge information. Therefore, a cross media knowledge information retrieval model using D-S evidence theory is proposed, which involves using approximate calculation methods to improve this theory for information fusion, reducing computational complexity, and using deep networks for fine-grained information retrieval to improve retrieval accuracy. The results showed that the improved theory enhanced computational efficiency by about 27.23 %. The memory usage was <60 %, and the average accuracy of information fusion reached 93.14 %. It also exhibited high recall and low false alarm rates. The cross media knowledge information retrieval model proposed in the study achieved accuracy of 92.64 %, 96.49 %, and 97.46 % on the three datasets used in the experiment, respectively. The study provides an effective, computationally efficient, and highly accurate model for cross media knowledge information retrieval, which is expected to promote research and application in this field. The combination of improved D-S evidence theory and deep networks provides a powerful approach to solving the problem of cross media heterogeneous information retrieval, which has a positive promoting effect on the processing and utilization of information in the information society.

Predictive Maintenance Based on Identity Resolution and Transformers in IIoT

The burgeoning development of next-generation technologies, especially the Industrial Internet of Things (IIoT), has heightened interest in predictive maintenance (PdM). Accurate failure forecasting and prompt responses to downtime are essential for improving the industrial efficiency. Traditional PdM methods often suffer from high false alarm rates and inefficiencies in complex environments. This paper introduces a predictive maintenance framework using identity resolution and a transformer model. Devices receive unique IDs via distributed identifiers (DIDs), followed by a state awareness model to assess device health from sensor signals. A sequence prediction model forecasts future signal sequences, which are then used with the state awareness model to determine future health statuses. Combining these predictions with unique IDs allows for the rapid identification of facilities needing maintenance. Experimental results show superior performance, with 99% accuracy for the state awareness model and a mean absolute error (MAE) of 0.062 for the sequence prediction model, underscoring the effectiveness of the framework.

Research on Network Intrusion Detection Technology Based on Artificial Intelligence

As network intrusion becomes more and more complicated and network characteristics become more and more diverse, the traditional network intrusion detection technology has some problems such as high false alarm rate, poor adaptability and low detection rate. The article provides an in-depth analysis of network intrusion detection techniques based on artificial intelligence to better solve these problems. In recent years, we have been witnessing great progress in image recognition, speech recognition, and natural language processing. The powerful advantage of artificial intelligence in processing large-scale and complex data provides new ideas for the processing of multi-attribute intrusion data. The introduction of AI technology in network intrusion detection can effectively improve the detection accuracy and reduce the false alarm rate and omission rate. Based on this, the article analyses the artificial intelligence network intrusion detection technology and discusses the network security management based on artificial intelligence for the reference of related personnel.

Review of intrusion detection system in cyber‐physical system based networks: Characteristics, industrial protocols, attacks, data sets and challenges

AbstractCyber‐Physical Systems (CPSs) provide critical infrastructure for the betterment of human lives thereby integrating cyber and physical components but the fusion of physical and digital components leads to an increase in the attack surface, which in turn provides opportunities for the attackers to intrude on these systems, which can affect the critical services like health care, water treatment facility, the electrical grid, hydropower plant, and so forth. The existing intrusion detection systems (IDSs) in CPSs are facing issues like poor detection accuracy, high false alarm rate and more computation time. Also, existing intrusion detection systems cannot identify new attacks that is, zero‐day assaults. Prerequisite exists for the design of the framework for detecting intrusions using artificial intelligence inspired approaches grounded on the principle of ensemble techniques. Intrusion detection framework assimilating ensemble‐based technique for CPS has been proposed. Since the proposed framework incorporates privileges of multiple techniques for intrusion detection and classification, hence the proposed framework may overcome the limitations of existing IDSs for CPS. Overview of cutting‐edge incursion identification methods for CPSs is presented. Several characteristics of CPS, industrial protocols, and anomaly detection techniques for intrusion detection are analyzed. Taxonomy of IDS for CPS has been proposed, taxonomy of attacks and threats on CPS has been intended. Research challenges for IDS in CPS are also uncovered from this review.

Cascaded-Filter-Based Reverberation Suppression Method of Short-Pulse Continuous Wave for Active Sonar

Reverberation is the main background interference in active sonar and seriously interferes with the extraction of the target echo. Active sonar systems can use short-pulse continuous wave (CW) signals to reduce the reverberation intensity. However, as the pulse width of the CW signals decreases, the reverberation envelope exhibits a high-frequency oscillating phenomenon. Active sonar often uses the cell average constant false alarm ratio (CA-CFAR) method to process the reverberation, which steadily decays with transmission distance. However, the high-frequency oscillation of the reverberation envelope deteriorates the performance of CA-CFAR, which causes a higher false alarm rate. To tackle this problem, the formation mechanism of the high-frequency oscillation characteristics of the reverberation envelope of the short-pulse-width CW signals is modeled and analyzed, and on this basis, an α filter is designed to suppress the high-frequency oscillation of the reverberation envelope before applying CA-CFAR. The simulation and lake trial results indicate that this method can effectively suppress high-frequency oscillations of the reverberation envelope, as well as exhibit robustness and resistance to reverberation interference.