Continuous Data Streams Research Articles

Diversity-Representativeness Replay and Knowledge Alignment for Lifelong Vehicle Re-identification

Lifelong vehicle re-identification (LVReID) aims to match a target vehicle across multiple cameras, considering non-stationary and continuous data streams, which fits the needs of the practical application better than traditional vehicle re-identification. Nonetheless, this area has received relatively little attention. Recently, methods for lifelong person re-identification (LPReID) have been emerging, with replay-based methods achieving the best results by storing a small number of instances from previous tasks for retraining, thus effectively reducing catastrophic forgetting. However, these methods cannot be directly applied to LVReID because they fail to simultaneously consider the diversity and representativeness of replayed data, resulting in biases between the subset stored in the memory buffer and the original data. They randomly sample classes, which may not adequately represent the distribution of the original data. Additionally, these methods fail to consider the rich variation in instances of the same vehicle class due to factors such as vehicle orientation and lighting conditions. Therefore, preserving more informative classes and instances for replay helps maintain information from previous tasks and may mitigate the model’s forgetting of old knowledge. In view of this, we propose a novel Diversity-Representativeness Dual-Stage Sampling Replay (DDSR) strategy for LVReID that constructs an effective memory buffer through two stages, i.e. , Cluster-Centric Class Selection and Diverse Instance Mining. Specifically, we first perform class-level sampling based on density in the clustered class-centered feature space and then further mine the diverse, high-quality instances within the selected classes. In addition, we introduce Maximum Mean Discrepancy loss to align the feature distribution between replay data and the new arrivals and apply L2 regularisation in the parameter space to facilitate knowledge transfer, thus enhancing the model’s generalization ability to new tasks. Extensive experiments demonstrate effective improvements of our method compared to current state-of-the-art lifelong ReID methods on the VeRi-776, VehicleID, and VERI-Wild datasets.

Sensitivity analysis for multi-measurement points based SHM in the mooring lines of floating offshore wind turbines

Abstract Structural health monitoring in floating offshore wind turbines’ mooring lines is vital for detecting early faults and preventing disruptions. Currently, sporadic and expensive monitoring is conducted via remote-operating vehicles. Efficient, automated, economical monitoring methods based on a continuous stream of data acquired via multiple measurement points in the wind turbine, are required. Such methods based on vibration signals have been investigated limitedly for steel chains. Recently, faulty synthetic mooring lines of a simulated 10MW semi-submersible wind turbine have been detected under varying environmental conditions and via the Functional Model Based Method (FMBM) equipped with functional models. The uncertainty due to the conditions’ stochasticity has been addressed by training the models using ten signal realizations per condition (wind) under the healthy wind turbine and from each of two measurement points. The current study presents a preliminary sensitivity analysis of the FMBM’s capability in detecting the simulated wind turbine’s faulty mooring lines when the functional models are trained using one signal realization per varying condition (wind) from each measurement point. The method’s effectiveness is evaluated with acceleration signals from 11 healthy/ 66 faulty (reduced stifness in one mooring line) cases. The FMBM detects all cases even when trained using one signal realization.

Dynamics of real-time forecasting failure and recovery due to data gaps: A study using EnKF-based assimilation with the Lorenz model

Data assimilation-based real-time forecasting is widely used in meteorological and hydrological applications, where continuous data streams are employed to update forecasts and maintain accuracy. However, the reliability of the data source can be compromised due to sensor and communication failures or physical or cyber-attacks, and the impact of data stream failures on the accuracy of the forecasting system is not well understood. This study aims to systematically investigate the process of data stream failure and recovery for the first time. To achieve this, data gaps with varying lengths and timings are introduced to EnKF-based data assimilation system on the Lorenz model operating in both chaotic and periodic modes. Results show that the forecasting error grows exponentially in the chaotic mode but was limited in the periodic mode from the start of the data gap. For chaotic mode, the recovery of the system depends on the length of the data gap if the model error is not saturated; after saturation, the timing of the data stream recovery is important. Moreover, even long after restarting the data assimilation in the chaotic mode, the forecasting system cannot fully restore the original accuracy, while the periodic mode is generally resilient to disruption. This research introduces new metrics for quantifying system resilience and provides crucial insights into the long-term implications of data gaps, advancing our understanding of forecasting system behavior and reliability.

Protecting Infinite Data Streams from Wearable Devices with Local Differential Privacy Techniques

The real-time data collected by wearable devices enables personalized health management and supports public health monitoring. However, sharing these data with third-party organizations introduces significant privacy risks. As a result, protecting and securely sharing wearable device data has become a critical concern. This paper proposes a local differential privacy-preserving algorithm designed for continuous data streams generated by wearable devices. Initially, the data stream is sampled at key points to avoid prematurely exhausting the privacy budget. Then, an adaptive allocation of the privacy budget at these points enhances privacy protection for sensitive data. Additionally, the optimized square wave (SW) mechanism introduces perturbations to the sampled points. Afterward, the Kalman filter algorithm is applied to maintain data flow patterns and reduce prediction errors. Experimental validation using two real datasets demonstrates that, under comparable conditions, this approach provides higher data availability than existing privacy protection methods for continuous data streams.

Learning incremental audio–visual representation for continual multimodal understanding

Deep learning methods have demonstrated remarkable success in processing static datasets for various video tasks. However, when confronted with continuous data streams, these approaches often encounter the challenge of catastrophic forgetting. This phenomenon leads to a significant decline in overall performance when learning new classes incrementally. Moreover, existing methods tend to overlook the correlation between audio and visual modalities in video incremental learning, despite their joint significance in scene comprehension. How to continuously learn from new classes while maintaining the knowledge of old videos with limited storage and computing resources is becoming imperative in the field of multimodal learning. In this paper, we introduce CavRL, a pioneering benchmark for audio–visual representation learning under class incremental scenarios. To mitigate catastrophic forgetting, we propose a rehearsal-based training approach that leverages a small exemplar set from previous classes. Our approach constrains the memory buffer within strict storage limits, optimizing exemplar selection by learning correlative audio–visual representations. Additionally, we employ a distillation method to mitigate forgetting in a self-supervised manner. Evaluations on two prevalent multimodal tasks: audio–visual event classification and audio–visual speaker recognition, which demonstrate that CavRL outperforms existing state-of-the-art incremental learning methods across various settings. We anticipate that CavRL will significantly advance research in continual multimodal learning.

Floor-Level Occupancy Estimation of a Multi-Story Building Using Coarse Wi-Fi Data

In recent years, there has been an extraordinary increase in wireless capable devices and network infrastructure, which spawned a corresponding rise in data produced from the interactions of these technologies. Mobile devices constantly roam, leading to a perpetual dialog between a mobile device and wireless access points. This dialogue generates a continuous stream of device-specific data, including but not limited to a device’s media access control address, time of access, and received signal strength. Given the knowledge of the access point’s location and received signal strength, it is possible to infer the position of user devices and estimate their mobility and occupancy. This article presents two methods for accurately measuring floor-level occupancy in a multi-story building at Texas State University using coarse Wi-Fi log data. The first method employs a static filter, while the second incorporates user-role data and user location to create a dynamic filter. Quantitative methods are used to evaluate these filters against field-collected reference data and existing internal people-counting sensors. Our results demonstrate that the dynamic filter, leveraging variable thresholds, provides a more accurate estimation of occupancy compared to the fixed 5-min static filter which consistently overestimated occupancy. This research sheds light on the potential of dynamic filters derived from user-role data for precise floor-level occupancy estimations, with implications for various applications.

An optimized hybrid encryption framework for smart home healthcare: Ensuring data confidentiality and security

This study proposes an optimized hybrid encryption framework combining ECC-256r1 with AES-128 in EAX mode, tailored for smart home healthcare environments, and conducts a comprehensive investigation to validate its performance. Our framework addresses current limitations in securing sensitive health data and demonstrates resilience against emerging quantum computing threats. Through rigorous experimental evaluation, we show that the proposed configuration outperforms existing solutions by delivering unmatched security, processing speed, and energy efficiency. It employs a robust yet streamlined approach, meticulously designed to ensure simplicity and practicality, facilitating seamless integration into existing systems without imposing undue complexity. Our investigation affirms the framework's capability to resist common cybersecurity threats like MITM, replay, and Sybil attacks while proactively considering quantum resilience. The proposed method excels in processing speed (0.006 seconds for client and server) and energy efficiency (3.65W client, 95.4W server), offering a quantum-resistant security level comparable to AES-128. This represents a security-efficiency ratio of 21.33 bits per millisecond, a 25.6% improvement in client-side processing speed, and up to 44% reduction in server-side energy consumption compared to conventional RSA-2048 methods. These improvements enable real-time encryption of continuous health data streams in IoT environments, making it ideal for IoT devices where AES-128′s smaller footprint is advantageous. By prioritizing high-grade encryption alongside ease of use and implementation, the proposed framework presents a future-proof solution that anticipates the trajectory of cryptographic standards amid advancing quantum computing technologies, signifying a pivotal advancement in safeguarding IoT-driven healthcare data.

Advancements in health informatics for monitoring perioperative patient data

Advancements in health informatics are revolutionizing perioperative care by enhancing patient monitoring, improving clinical decision-making, and promoting patient safety. The integration of electronic health records (EHR) facilitates seamless information flow across all perioperative stages, ensuring up-to-date patient data is accessible to the entire healthcare team. This accessibility reduces errors, improves coordination, and enhances overall patient outcomes. EHR systems also support standardized documentation, reducing variability in record-keeping and ensuring comprehensive patient information is consistently captured. Real-time data analytics and predictive modeling are powerful tools in perioperative settings, enabling clinicians to anticipate and manage potential complications proactively. By analyzing vast datasets, predictive models identify patterns and risk factors associated with adverse outcomes, allowing for targeted interventions and personalized care plans. Real-time analytics provide continuous monitoring of patient status during surgery, offering immediate feedback and enabling prompt adjustments to ensure patient stability. Wearable devices and remote monitoring systems further augment real-time data analytics by continuously tracking physiological parameters throughout the surgical process and into the postoperative period. This continuous data stream allows for ongoing assessment of patient status and early detection of complications, enhancing patient safety and outcomes. Health informatics solutions, including automated alerts and decision support tools, minimize human error and ensure adherence to clinical guidelines. Automated reminders for critical actions, such as timely antibiotic administration, significantly reduce the incidence of surgical site infections. Decision support tools provide real-time, evidence-based recommendations at the point of care, supporting clinicians in making informed decisions and improving patient outcomes. Additionally, health informatics facilitates continuous quality improvement by enabling the collection and analysis of performance data, informing targeted interventions and process improvements. Overall, advancements in health informatics are transforming perioperative care, offering numerous benefits in terms of efficiency, accuracy, and patient safety, ultimately leading to better patient outcomes and higher quality of care. Continued innovation in this field will be essential for maintaining and improving the standards of perioperative care.

How Users Assess Privacy Risks in the Internet of Things: The Role of Framing, Comparing, and Educating

With the advent of the Internet of Things (IoT), it has become increasingly challenging for users to assess the privacy risks associated with consumer products and the continuous stream of user data needed to operate them. In this study, we propose and test three mechanisms with the potential to help users make more accurate assessments of privacy risks. We refer to these mechanisms as framing (i.e., presenting information on the collection and use of user data with or without direct reference to privacy risks), comparing (i.e., presenting a product and the associated information on data collection and use with or without reference to an alternative product), and educating (i.e., augmenting users’ general privacy literacy). To assess these mechanisms in different IoT contexts, we conducted two scenario-based online experiments with reference to a telematics device ( n = 317) and a fitness tracker ( n = 356). In both studies, we find that actual privacy risks as manipulated in the experiment are only moderately related to the privacy risks perceived by users. However, comparing and educating each helped users make more accurate privacy risk assessments. In Study 2, framing and comparing jointly enabled especially users with low privacy literacy to assess privacy risks more accurately. These findings have meaningful implications for key actors in the IoT ecosystem and those regulating it.

Stream-aware indexing for distributed inequality join processing

Inequality join is an operator to join data on inequality conditions and it is a fundamental building block for applications. While methods and optimizations exist for efficient inequality join in batch processing, little attention has been given to its streaming version, particularly to large-scale data-intensive applications that run on Distributed Stream Processing Systems (DSPSs). Designing an inequality join in streaming and distributed settings is not an easy task: (i) indexes have to be employed to efficiently support inequality-based comparisons, but the continuous stream of data imposes continuous insertions, updates, and deletions of elements in the indexes—hence a huge overhead for the DSPSs; (ii) oftentimes real data is skewed, which makes indexing even more challenging.To address these challenges, we propose the Stream-Aware inequality join (STA), an indexing method that can reduce redundancy and index update overhead. STA builds a separate in-memory index structure for hotkeys, i.e., the most frequently used keys, which are automatically identified with an efficient data sketch. On the other hand, the cold keys are treated using a linked set of index structures. In this way, STA avoids many superfluous index updates for frequent items. Finally, we implement four state-of-the-art inequality join solutions for a widely employed DSPS (Apache Storm) and compare their performance with STA on four real-world data sets and a synthetic one. The results of our experimental evaluation reveal that our stream-aware approach outperforms existing solutions.

IPREDICT: Characterization of Asthma Triggers and Selection of Digital Technology to Predict Changes in Disease Control.

The iPREDICT program aimed to develop an integrated digital health solution capable of continuous data streaming, predicting changes in asthma control, and enabling early intervention. As part of the iPREDICT program, asthma triggers were characterized by surveying 221patients (aged ≥18 years) with self-reported asthma for a risk-benefit analysis of parameters predictive of changes in disease control. Seventeen healthy volunteers (age25-65 years) tested 13 devices to measure these parameters and assessed their usability attributes. Patients identified irritants such as chemicals, allergens, weather changes, and physical activity as triggers that were the most relevant to deteriorating asthma control. Device testing in healthy volunteers revealed variable data formats/units and quality issues, such as missing data and low signal-to-noise ratio. Based on user preference and data capture validity, a spirometer, vital sign monitor, and sleep monitor formed the iPREDICT integrated system for continuous data streaming to develop a personalized/predictive algorithm for asthma control. These findings emphasize the need to systematically compare devices based on several parameters, including usability and data quality, to develop integrated digital technology programs for asthma care.

Federated Learning Empowered Real-Time Medical Data Processing Method for Smart Healthcare.

Computer-aided diagnosis (CAD) has always been an important research topic for applying artificial intelligence in smart healthcare. Sufficient medical data are one of the most critical factors in CAD research. However, medical data are usually obtained in chronological order and cannot be collected all at once, which poses difficulties for the application of deep learning technology in the medical field. The traditional batch learning method consumes considerable time and space resources for real-time medical data, and the incremental learning method often leads to catastrophic forgetting. To solve these problems, we propose a real-time medical data processing method based on federated learning. We divide the process into the model stage and the exemplar stage. In the model stage, we use the federated learning method to fuse the old and new models to mitigate the catastrophic forgetting problem of the new model. In the exemplar stage, we use the most representative exemplars selected from the old data to help the new model review the old knowledge, which further mitigates the catastrophic forgetting problem of the new model. We use this method to conduct experiments on a simulated medical real-time data stream. The experimental results show that our method can learn a disease diagnosis model from a continuous medical real-time data stream. As the amount of data increases, the performance of the disease diagnosis model continues to improve, and the catastrophic forgetting problem has been effectively mitigated. Compared with the traditional batch learning method, our method can significantly save time and space resources.

SS-CRE: A Continual Relation Extraction Method Through SimCSE-BERT and Static Relation Prototypes

Continual relation extraction aims to learn new relations from a continuous stream of data while avoiding forgetting old relations. Existing methods typically use the BERT encoder to obtain semantic embeddings, ignoring the fact that the vector representations suffer from anisotropy and uneven distribution. Furthermore, the relation prototypes are usually computed by memory samples directly, resulting in the model being overly sensitive to memory samples. To solve these problems, we propose a new continual relation extraction method. Firstly, we modified the basic structure of the sample encoder to generate uniformly distributed semantic embeddings using the supervised SimCSE-BERT to obtain richer sample information. Secondly, we introduced static relation prototypes and dynamically adjust their proportion with dynamic relation prototypes to adapt to the feature space. Lastly, through experimental analysis on the widely used FewRel and TACRED datasets, the results demonstrate that the proposed method effectively enhances semantic embeddings and relation prototypes, resulting in a further alleviation of catastrophic forgetting in the model. The code will be soon released at https://github.com/SuyueW/SS-CRE.

Using Graphs to Perform Effective Sensor-Based Human Activity Recognition in Smart Homes.

There has been a resurgence of applications focused on human activity recognition (HAR) in smart homes, especially in the field of ambient intelligence and assisted-living technologies. However, such applications present numerous significant challenges to any automated analysis system operating in the real world, such as variability, sparsity, and noise in sensor measurements. Although state-of-the-art HAR systems have made considerable strides in addressing some of these challenges, they suffer from a practical limitation: they require successful pre-segmentation of continuous sensor data streams prior to automated recognition, i.e., they assume that an oracle is present during deployment, and that it is capable of identifying time windows of interest across discrete sensor events. To overcome this limitation, we propose a novel graph-guided neural network approach that performs activity recognition by learning explicit co-firing relationships between sensors. We accomplish this by learning a more expressive graph structure representing the sensor network in a smart home in a data-driven manner. Our approach maps discrete input sensor measurements to a feature space through the application of attention mechanisms and hierarchical pooling of node embeddings. We demonstrate the effectiveness of our proposed approach by conducting several experiments on CASAS datasets, showing that the resulting graph-guided neural network outperforms the state-of-the-art method for HAR in smart homes across multiple datasets and by large margins. These results are promising because they push HAR for smart homes closer to real-world applications.

The Pulse of Progress: Smart Wearables and the Evolution of Cardiovascular Monitoring Technologies

This paper explores the transformative role of smart wearables in the evolution of cardiovascular monitoring technologies. With cardiovascular diseases (CVDs) as the leading cause of mortality globally, there's a critical need for continuous, real-time health monitoring outside traditional clinical settings. Smart wearables, ranging from fitness trackers to specialized health-monitoring devices, have emerged as pivotal tools in this regard. They offer personalized, preventive, and real-time management of health, leveraging advancements in technology to provide a continuous stream of health data previously inaccessible. This systematic review draws upon recent studies to examine the effectiveness of wearable technologies in detecting, predicting, and managing CVDs.

Optofluidic time-stretch imaging flow cytometry with a real-time storage rate beyond 5.9 GB/s.

Optofluidic time-stretch imaging flow cytometry (OTS-IFC) provides a suitable solution for high-precision cell analysis and high-sensitivity detection of rare cells due to its high-throughput and continuous image acquisition. However, transferring and storing continuous big data streams remains a challenge. In this study, we designed a high-speed streaming storage strategy to store OTS-IFC data in real-time, overcoming the imbalance between the fast generation speed in the data acquisition and processing subsystem and the comparatively slower storage speed in the transmission and storage subsystem. This strategy, utilizing an asynchronous buffer structure built on the producer-consumer model, optimizes memory usage for enhanced data throughput and stability. We evaluated the storage performance of the high-speed streaming storage strategy in ultra-large-scale blood cell imaging on a common commercial device. The experimental results show that it can provide a continuous data throughput of up to 5891 MB/s.

CLADSI: Deep Continual Learning for Alzheimer's Disease Stage Identification Using Accelerometer Data.

Alzheimer's disease (AD) is a neurodegenerative disorder that can cause a significant impairment in physical and cognitive functions. Gait disturbances are also reported as a symptom of AD. Previous works have used Convolutional Neural Networks (CNNs) to analyze data provided by motion sensors that monitor Alzheimer's patients. However, these works have not explored continual learning algorithms that allow the CNN to configure itself as it receives new data from these sensors. This work proposes a method aimed at enabling CNNs to learn from a continuous stream of data from motion sensors without having full access to previous data. The CNN identifies the stage of AD from the analysis of data provided by motion sensors. The work includes an experimentation with data captured by accelerometers that monitored the activity of 35 Alzheimer's patients for a week in a daycare center. The CNN achieves an accuracy of 86,94%, 86,48% and 84,37% for 2, 3 and 4 experiences respectively. The proposal provides advantages to working with a continuous stream of data so that the CNN are constantly self-configuring without the intervention of a human. The work can be considered as promising and helpful in finding deep learning solutions in medical cases in which patients are constantly monitored.

Human activity recognition using a single-photon direct time-of-flight sensor

Single-Photon Avalanche Diode (SPAD) direct Time-of-Flight (dToF) sensors provide depth imaging over long distances, enabling the detection of objects even in the absence of contrast in colour or texture. However, distant objects are represented by just a few pixels and are subject to noise from solar interference, limiting the applicability of existing computer vision techniques for high-level scene interpretation. We present a new SPAD-based vision system for human activity recognition, based on convolutional and recurrent neural networks, which is trained entirely on synthetic data. In tests using real data from a 64×32 pixel SPAD, captured over a distance of 40 m, the scheme successfully overcomes the limited transverse resolution (in which human limbs are approximately one pixel across), achieving an average accuracy of 89% in distinguishing between seven different activities. The approach analyses continuous streams of video-rate depth data at a maximal rate of 66 FPS when executed on a GPU, making it well-suited for real-time applications such as surveillance or situational awareness in autonomous systems.

B-Detection: Runtime Reliability Anomaly Detection for MEC Services With Boosting LSTM Autoencoder

By pushing computing resources from the cloud to the network edge close to mobile users, mobile edge computing (MEC) enables low latency for a wide variety of applications. Nevertheless, in dynamic MEC systems, MEC services are challenged by the risks of runtime reliability anomalies. Detecting runtime reliability anomalies for MEC services is challenging yet critical to ensuring the stability of MEC systems. The effectiveness of existing anomaly detection methods suffers from poor performance when handling MEC services' large-volume, continuous, and volatile reliability streaming data. The key is to identify significant changes in the distribution of MEC services' current reliability streaming data compared with their historical performance. Inspired by concept drift, this paper proposes B-Detection, a boosting Long Short-Term Memory (LSTM) Autoencoder for detecting MEC services' runtime reliability anomalies based on distribution dissimilarity evaluation. B-Detection employs a deep learning method named LSTM Autoencoder to characterize the MEC services' historical reliability data distribution. To cope with the challenge of modeling complex distribution characteristics of MEC services' historical reliability streaming data and guarantee the real-time performance of B-Detection, we enhance LSTM Autoencoder with a weight-based reservoir sampling technique and an LSTM boosting algorithm. The reconstruction loss of the trained LSTM Autoencoder model is estimated for the up-to-date reliability streaming data, and the result is used to infer MEC services' runtime reliability anomalies. The performance of B-Detection is verified through a series of experiments conducted on a real-world dataset.

Revolutionizing Cardiovascular Health: A Machine Learning Approach for Predictive Analysis and Personalized Intervention in Heart Disease

Abstract: Cardiovascular Diseases (CVDs) continue to be a leading cause of global morbidity and mortality, necessitating innovative approaches for early detection and personalized interventions. This research explores the transformative potential of machine learning (ML) in revolutionizing cardiovascular health. Leveraging advanced predictive analytics, our study employs a comprehensive dataset of cardiovascular parameters, incorporating clinical, genetic, and lifestyle factors. The machine learning model developed demonstrates remarkable accuracy in predicting the risk of heart disease, enabling early identification of individuals susceptible to CVD. This predictive analysis empowers healthcare professionals with a powerful tool for pre-emptive intervention and tailored treatment strategies. By considering individual variations in genetic predispositions, lifestyle choices, and clinical data, our approach moves beyond traditional risk assessment models, paving the way for a more personalized and effective healthcare paradigm. Furthermore, the integration of real-time monitoring devices and continuous data streams enhances the adaptability and responsiveness of our model. This allows for dynamic adjustments in treatment plans, ensuring ongoing optimization based on the evolving health status of each patient. The synergy between machine learning and cardiovascular health not only augments diagnostic precision but also facilitates a proactive healthcare ecosystem that prioritizes preventive measures.