Speech overlap, which occurs when multiple people speak simultaneously, poses a significant challenge in audio and speech processing. The presence of overlapping speech segments significantly degrades the performance of technologies such as Automatic Speech Recognition (ASR), speaker identification, and diarization systems. This degradation in performance becomes more significant in diverse acoustic environments with background noise and reverberation. To effectively address this issue, we introduce BiConNet. This novel dual-branch architecture combines the strengths of Convolutional Neural Networks (CNN) and Bidirectional Long Short-Term Memory (BiLSTM) for robust detection of overlapping speech in diverse acoustic conditions. The CNN branch is used for frame-level spectral feature extraction, while the BiLSTM branch captures temporal dependencies from both forward and backward directions. Features from both branches are concatenated, resulting in a robust feature representation. We also examined the impact of Mel Frequency Cepstral Coefficients (MFCC), Gammatone Frequency Cepstral Coefficients (GTCC), and Power Normalized Cepstral Coefficients (PNCC) as spectral-based features on BiConNet's performance. To validate its effectiveness in various acoustic environments, we developed a constructed data set derived from the GRID corpus, including conversations with different gender combinations and recording conditions, such as clean, noisy, reverberant, and combined noise and reverberation conditions. Experimental results show that BiConNet outperforms various state-of-the-art methods in detecting overlapping speech segments under these conditions. Furthermore, our analysis of computational efficiency reveals that BiConNet provides competitive training and inference times, demonstrating its practicability for real-world applications.

People can often acquire knowledge dynamically and rapidly from different types of data, yet existing incremental learning algorithms are still computationally time consuming and most of stream learning methods are mainly designed for streaming data while ignoring other types of data. Hence, this paper proposes a novel dynamic concept learning (CL) algorithm by imitating human cognitive learning processes from the perspective of brain logical cognition, which is named stream concept-cognitive computing system (streamC3S). For streamC3S, it mainly consists of three aspects: the concept space, CL process, and model update process. Moreover, considering the concept drift frequently occurs in the streaming data over time, an extended version of streamC3S (namely, streamC3SE) is also proposed in this work. Specifically, we first show the related theories for streamC3S and streamC3SE on the basis of the concept space. Then an overall framework and its corresponding algorithm are shown. Finally, experimental results on various types of datasets, including the standard machine learning datasets, streaming datasets, image datasets, and two traffic data streams, validate the effectiveness of our streamC3S and streamC3SE compared to the state-of-the-art incremental learning and stream learning algorithms.

A document by Muhammad Farhan Shahid . Click on the document to view its contents.

Most known quantum magnetometers — proton, Overhauser, many types of sensors with optical pumping — have a periodic signal at the output. Its frequency is directly proportional to the magnitude of the magnetic field with high accuracy. Thus, by measuring this frequency, we can obtain the value of the magnetic field. In many applications, such as aeromagnetic measurements, we need to increase the sampling rate, but at the same time we reduce the accuracy of the measurements. The article aims to obtain the optimal frequency measurement algorithm depending on the characteristics of the equipment. We will also describe how to use a satellite navigation system as a time reference to ensure accuracy. All ideas are illustrated by experiments with various optically pumped cesium magnetometers. The reference standard frequency of the frequency counter used is 1 GHz. It is shown that for this frequency counter a sampling rate of up to 1 kHz can be obtained with any of the considered sensors.

Recommender Systems are information systems that provide personalized suggestions or recommendations to users based on their preferences, interaction history, and behavior patterns. An important use of recommender systems is in Virtual Learning Environments (VLEs), i.e. remote learning platforms where students can access educational materials, activities, assessments, etc., where they offer significant benefits to both students and instructors. Such benefits include personalized learning, increased engagement and effectiveness of both teaching and learning tasks. Challenges for the use of recommendation systems in VLEs include the cold start problem, data sparsity, and limited coverage. To deal with these challenges, we propose G-Learn, a recommendation system for contents in VLEs that can operate in both supervised and unsupervised modes. The recommendations are based on graph machine learning techniques combined with keyword mining and similarity, allowing the system to recommend educational materials that are adapted to the performance of each student detected when the student solves questions on the platform. We show the effectiveness of G-learn in a realistic scenario, using data from the  VLE Homero, a computer science platform supported by the federal government of Brazil. We compare different settings for the keyword mining and machine learning techniques. Real data from students is used to validate the recommendation graphs, where G-learn obtained an average f1-score of 0.64 in the unsupervised mode and 0.95 in the supervised one.

This paper proposes a dynamic event-triggered control strategy for the leader-following multi-agent control under directed topology.  A synthesis approach combining distributed controllers and observers design is developed under a dynamic sampling scheme, and only local information is required for each agent to implement the proposed method. The control protocol incorporates model-based estimation and clock-like auxiliary dynamic variables to prolong the inter-event time as long as possible.  Sufficient conditions for leader-following consensus control are established by linear matrix inequalities, and an explicit inter-event time is given to enable flexible tuning. Due to the carefully selected Lyapunov function, the proposed method exhibits significant advantages over the dynamic event-triggered control methods described in the existing literature. Compared to the existing static event-triggered strategy, the proposed approach significantly reduces the utilization of communication resources while preserving asymptotic convergence to the state consensus.  The validity and effectiveness of the proposed theoretical results are demonstrated by comparative simulations.

A document by Alexander Franke . Click on the document to view its contents.

The electromagnetic fields of point sources with time varying charges moving in the vacuum are derived using the Liénard-Wiechert potentials. The properties of the propagation velocities and the Doppler effect are discussed based on their far fields. The results show that the velocity of the electromagnetic waves and the velocity of the sources cannot be added like vectors; the velocity of electromagnetic waves of moving sources are anisotropic in the vacuum; the transverse Doppler shift is intrinsically included in the fields of the moving sources and is not a pure relativity effect caused by time dilation. Since the fields are rigorous solutions of the Maxwell's equations, the findings can help us to abort the long-standing misinterpretations concerning about the classic mechanics and the classic electromagnetic theory. Although it may violate the theory of the special relativity, we show mathematically that, when the sources move faster than the light in the vacuum, the electromagnetic barriers and the electromagnetic shock waves can be clearly predicted using the exact solutions. Since they cannot be detected by observers in the region outside their shock wave zones, an intuitive and reasonable hypothesis can be made that the superluminal sources may be considered as a kind of electromagnetic blackholes.

There is great interest in alternative pollination strategies for crop production in the face of climate change and perennial threats to the traditional pollination mechanisms. This review explores the potential for robotic pollination in response to these challenges to crop fertilization and global food production. Herein we describe the viability of novel robotic systems equipped with artificial intelligence and machine vision, as alternatives to traditional insect pollination. We examine the technological progress and challenges for both aerial and ground-based robotic artificial pollination systems and emphasize the need for continued research and development in this area to ensure sustainable agricultural productivity. This paper highlights the importance of robotic pollination as a practical and environmentally sustainable approach in modern agriculture, amidst burgeoning ecological threats and a dwindling agricultural workforce.

As a machine learning approach, domain adaptation methods are widely applied in cross-scenario fault diagnosis. However, the target domain may need more annotated data, posing challenges to the performance of domain adaptation methods.

This paper discusses the issues and challenges that can be found in data communication and networking, as well as possible solutions to those issues

Technology, including the internet, phones, computers, tablets, platforms, networks, software applications, and databases, has significantly impacted our daily lives. The integration of e-commerce platforms in rural agricultural sectors has opened new opportunities for farmers and producers to reach wider markets. These platforms can facilitate growth, efficiency, and market access for rural farmers. As rural revitalization progresses, e-commerce can help agriculture development through diversification, intelligent agriculture, and e-commerce trading platforms. These platforms have the potential to revolutionize traditional farming by creating direct links between farmers and buyers. This paper aims to study the development strategy of ecommerce in rural areas, focusing on marketing status, problems, and specific development strategies, to promote the development of e-commerce in rural areas under the background of rural revitalization.

A single-event transient (SET) tolerant dynamic bias comparator leveraging path-splitting is presented in this paper. The dynamic bias structure with a tail capacitor is found to be vulnerable to positive and negative SET transients. A negative transient on the tail transistor doubles the energy per conversion. A positive transient on the tail MOS capacitor reduces the preamplifier gain by 35.7%, which increases the metastability error rate in the subsequent latch. A path splitting technique is applied to mitigate the positive SET. Simulations in a 65-nm CMOS process showed that a two path split can recover the comparator output voltage swing by 12.7%. Compared with the state-of-the-art CMOS radiation-hardened by design (RHBD) dynamic comparators, this circuit achieved the lowest energy per comparison while maintaining low input referred noise and gaining SET tolerance in the evaluation phase.

This paper describes the process followed to implement a system to characterize the complex permittivity of materials in the 10-330 GHz frequency band. Firstly, the method used and the system's calibration process are shown, consisting of a double calibration TRL (Thru-Reflect-Line) and GRL (Gated-Reflect-Line). Subsequently, a smoothing technique is used to improve the accuracy of the results. Finally, a test is performed on quartz and glass fiber samples, showing that the results are quite reliable over the entire measured bandwidth.

Age-related cognitive impairment is generally characterized by gradual memory loss and decision-making difficulties. The aim of this study is to investigate multi-level support and suggest relevant helping means for the elderly with mild cognitive impairment as well as their caregivers as the primary end-users. This work reports preliminary results on an intelligent assistive system achieved through the integration of the Internet of Things, augmented reality, and adaptive fuzzy decision-making methods. The proposed system operates in different modes, including automated and semi-automated modes. The former helps the user complete their daily life activities by showing augmented reality messages or making automatic changes; while the latter allows manual changes after the real-time assessment of the user’s cognitive state based on the augmented reality serious game score. We have also evaluated the accuracy of the serious game score with 37 elderly participants and compared it with users’ paper-based cognitive test results. We further noted that there is an acceptable correlation between the paper-based test and users’ serious game scores. Moreover, we observed that the system response in

Rising shares of renewable generation raises uncertainty and thus the number of possible power flow scenarios in the power system, which in turn increases the possibility for unforeseen contingencies, such as power line or generator failures and their combinations. Therefore, operators cannot longer rely only on operational experience to deal with every contingency. Our proposed method involves identifying the most effective countermeasures to minimize the impact of contingencies on the power system. We take into account various options, such as load shedding, adjusting phase-shifting transformer angles, and injecting active power using fast elements. The proposed approach considers primary control of generators and its limitations in order to compensate for power imbalances in the system. The problem is formulated as a mixed-integer linear optimization problem, employing DC power flow equations. The applicability of the approach is evaluated on the IEEE 39-bus system, and the scalability of the approach is shown on five systems with up to 6470 buses.

Nowadays, special focus on intelligent and context-aware data management is given in the intersection between Internet of Things (IoT) and Edge Computing (EC), where humongous volumes of data are collected and transferred towards loud data centers. Edge nodes act as intermediaries between IoT devices and data centers, thus, their behaviour and performance are critical for supporting high-quality, data-driven pervasive services to end-users. In this paper, we propose a context-aware strategy for data selectivity at the network edge that deals with the challenges of limited computational and storage resources. Our goal is to efficiently learn the access patterns of incoming data-driven tasks and predict which data are relevant, thus, should be stored in nodes’ local datasets. Tasks are associated with processing activities over local datasets directly indicating the data sub-spaces required to be accessed. We adopt one-class classification statistical learning to identify the data sub-spaces requested by tasks while eliminating outlier requests. In turn, we develop an uncertainty-driven mechanism based on fuzzy inference rules that determines a data filter mechanism capable of predicting the future relevant data sub-spaces. We analytically describe our mechanism and comprehensively evaluate and compare it against baselines and models found in the literature showcasing its applicability in pervasive edge computing environments.

Advances in Internet of Things (IoT) technologies have resulted in a significant surge in the utilization of sensor devices across diverse domains for environmental sensing and monitoring. The applications of IoT sensor devices in environmental monitoring span a wide range, including the surveillance of biodiverse areas such as peatlands, forests, and oceans, as well as air quality monitoring, commercial agriculture, and the safeguarding of endangered species. This paper provides a long term evaluation of IoT sensors data quality in environmental monitoring networks, particularly focusing on peatland regions. IoT sensors have the capacity to provide high resolution spatiotemporal dataset in environmental monitoring networks. Sensor data quality plays significant role in increasing the adoption of IoT devices for environmental data gathering. However, due to the nature of deployment (i.e., in harsh and unfavourable weather conditions), coupled with the limitations of low-cost components, IoT sensors are prone to collection of erroneous data, also the nature of peatland ecosystems presents unique challenges in data quality assurance due to their complex and dynamic characteristics. This paper identifies specific challenges and issues related to IoT sensor data quality in different peatland ecotopes. These challenges include sensor placement and calibration, data validation and fusion, environmental interference, and the management of data gaps and uncertainties. To address these challenges, the paper presents and evaluates methods for improving data quality in peatland monitoring networks. These methods encompass advanced sensor calibration techniques, data validation algorithms, machine learning approaches, data processing and data fusion strategies.