Human activity recognition(HAR)is a method to predict human activities from sensor signals using machine learning(ML)techniques.HAR systems have several applications in various domains,including medicine,surveillance,...Human activity recognition(HAR)is a method to predict human activities from sensor signals using machine learning(ML)techniques.HAR systems have several applications in various domains,including medicine,surveillance,behavioral monitoring,and posture analysis.Extraction of suitable information from sensor data is an important part of the HAR process to recognize activities accurately.Several research studies on HAR have utilizedMel frequency cepstral coefficients(MFCCs)because of their effectiveness in capturing the periodic pattern of sensor signals.However,existing MFCC-based approaches often fail to capture sufficient temporal variability,which limits their ability to distinguish between complex or imbalanced activity classes robustly.To address this gap,this study proposes a feature fusion strategy that merges time-based and MFCC features(MFCCT)to enhance activity representation.The merged features were fed to a convolutional neural network(CNN)integrated with long shortterm memory(LSTM)—DeepConvLSTM to construct the HAR model.The MFCCT features with DeepConvLSTM achieved better performance as compared to MFCCs and time-based features on PAMAP2,UCI-HAR,and WISDM by obtaining an accuracy of 97%,98%,and 97%,respectively.In addition,DeepConvLSTM outperformed the deep learning(DL)algorithms that have recently been employed in HAR.These results confirm that the proposed hybrid features are not only practical but also generalizable,making them applicable across diverse HAR datasets for accurate activity classification.展开更多
The concept of Human Activity Recognition(HAR)is integral to applications based on Internet of Things(IoT)-enabled devices,particularly in healthcare,fitness tracking,and smart environments.The streams of data from we...The concept of Human Activity Recognition(HAR)is integral to applications based on Internet of Things(IoT)-enabled devices,particularly in healthcare,fitness tracking,and smart environments.The streams of data from wearable sensors are rich in information,yet their high dimensionality and variability pose a significant challenge to proper classification.To address this problem,this paper proposes hybrid architectures that integrate traditional machine learning models with a deep neural network(DNN)to deliver improved performance and enhanced capabilities for HAR tasks.Multi-sensor HAR data were used to systematically test several hybrid models,including:RF+DNN(Random Forest+Deep Neural Network),XGB+DNN(XGBoost+DNN),GB+DNN(Gradient Boosting+DNN),KNN+DNN(K-Nearest Neighbors+DNN),and DT+DNN(Decision Tree+DNN).The RF+DNN model was the most accurate,achieving a 97.03%score with excellent precision,recall,and F1-score.These findings demonstrate that hybrid machine learning and deep learning systems have a promising future in IoT-based HAR applications.The model provides a novel solution for developing smart and trustworthy monitoring systems that support real-time analytics,patient surveillance,and other IoT applications.展开更多
The Bayan Har block,one of China's most seismically active regions,has experienced multiple major earthquakes(≥M 7.0)in recent years.It is a key area for investigating the interactions between the Qinghai-Xizang(...The Bayan Har block,one of China's most seismically active regions,has experienced multiple major earthquakes(≥M 7.0)in recent years.It is a key area for investigating the interactions between the Qinghai-Xizang(Qingzang)Plateau and adjacent blocks,plateau uplift,and strong earthquake mechanisms.P-wave velocity and crustal composition provide key constraints on the properties of distinct tectonic units and their evolutionary modification processes.Based on the results of 8 Deep Seismic Sounding(DSS)profiles completed in the Bayan Har block and surrounding areas over the past 20 years,We constructed one-dimensional P-wave velocity models for the crust of Bayan Har block,Qilian fold belt,Qinling fold belt,Alxa block,Ordos block and Sichuan basin.Furthermore,crustal composition models for different tectonic units were established based on these results.The results reveal that the crustal thickness of the Bayan Har block gradually decreases towards the NNE,NE,and SE directions,while the average crustal velocity increases correspondingly.The felsic layer in the crust accounts for more than half of the total crustal thickness.The mafic content within the crust of different tectonic units exhibits notable variations,which may reflect that the Bayan Har block,Qilian fold belt,and Qinling fold belt have experienced more intensive lithospheric evolution processes compared to Ordos basin and Sichuan basin.The seismicity distribution in this region is significantly controlled by crustal velocity and composition heterogeneity across the Bayan Har block and adjacent areas,which demonstrates that earthquakes within and around the Bayan Har block exhibit both high frequency and larger magnitudes.These seismic characteristics primarily result from intense crustal stress accumulation and release during the outward expansion of the Qingzang Plateau.展开更多
The ISO Annual Meeting is the world’s premier event for the international standards community.This year’s event is hosted by RSB,the national standards body of Rwanda.The theme“United for impact”calls us to harnes...The ISO Annual Meeting is the world’s premier event for the international standards community.This year’s event is hosted by RSB,the national standards body of Rwanda.The theme“United for impact”calls us to harness our collective strength in a world that needs bold,lasting change.The ISO Annual Meeting convenes global leaders and change-makers to explore how international standards can unlock progress,foster trust and drive lasting solutions to our shared challenges.展开更多
The increase in wood and wood-based products in the construction and furniture sectors has grown exponentially,generating severe environmental and socioeconomic impacts.Particleboard panels have been the main cost-ben...The increase in wood and wood-based products in the construction and furniture sectors has grown exponentially,generating severe environmental and socioeconomic impacts.Particleboard panels have been the main cost-benefit option on the market due to their lightness and lower cost compared to solid wood.However,the synthetic adhesives used in producing traditional particleboard panels cause serious harm to human health.Developing particleboard panels with fibrous waste and natural adhesives could be a sustainable alternative for these sectors.The work aimed to create particleboards with fibrous wastes from the pseudostem of the banana tree(Musa paradisiaca)and different proportions of the natural adhesive cassava starch-CS in replacement of synthetic adhesive urea-formaldehyde-UF.Five experimental groups were manufactured with banana trees and different percentages of UF and CS adhesives,namely(100UF–0%CS),(50%UF–50%CS),(30%UF–70%CS),(10%UF–90%CS)and(0%UF–100%CS).The particleboards had their physical-mechanical properties determined.The apparent density values did not show significant variation between the assessed treatments.Regarding the water absorption and thickness swelling,the best performances were observed for the panels made without the addition of CS(100%UF).For the mechanical properties of static bending strength and Janka hardness,it was identified that adding up to 50%CS did not interfere with the quality of the panels.These analyses show that the particleboard panels produced with wastes of the banana tree bonded with natural CS adhesivemay be an economically viable and environmentally correct alternative,positively strengthening the development of sustainable strategies.展开更多
Human Activity Recognition(HAR)represents a rapidly advancing research domain,propelled by continuous developments in sensor technologies and the Internet of Things(IoT).Deep learning has become the dominant paradigm ...Human Activity Recognition(HAR)represents a rapidly advancing research domain,propelled by continuous developments in sensor technologies and the Internet of Things(IoT).Deep learning has become the dominant paradigm in sensor-based HAR systems,offering significant advantages over traditional machine learning methods by eliminating manual feature extraction,enhancing recognition accuracy for complex activities,and enabling the exploitation of unlabeled data through generative models.This paper provides a comprehensive review of recent advancements and emerging trends in deep learning models developed for sensor-based human activity recognition(HAR)systems.We begin with an overview of fundamental HAR concepts in sensor-driven contexts,followed by a systematic categorization and summary of existing research.Our survey encompasses a wide range of deep learning approaches,including Multi-Layer Perceptrons(MLP),Convolutional Neural Networks(CNN),Recurrent Neural Networks(RNN),Long Short-Term Memory networks(LSTM),Gated Recurrent Units(GRU),Transformers,Deep Belief Networks(DBN),and hybrid architectures.A comparative evaluation of these models is provided,highlighting their performance,architectural complexity,and contributions to the field.Beyond Centralized deep learning models,we examine the role of Federated Learning(FL)in HAR,highlighting current applications and research directions.Finally,we discuss the growing importance of Explainable Artificial Intelligence(XAI)in sensor-based HAR,reviewing recent studies that integrate interpretability methods to enhance transparency and trustworthiness in deep learning-based HAR systems.展开更多
Male sterile lines serve as a pivotal tool in plant breeding,offering an effective strategy for developing hybrid cultivars.In cucumber(Cucumis sativus L.),the creation of male sterile lines,most of which exhibit abno...Male sterile lines serve as a pivotal tool in plant breeding,offering an effective strategy for developing hybrid cultivars.In cucumber(Cucumis sativus L.),the creation of male sterile lines,most of which exhibit abnormal pollen development,is essential for harnessing heterosis.Accumulating evidence suggests that male sterility in plants is linked to the overaccumulation of Reactive Oxygen Species(ROS),which causes oxidative damage to biomolecules and cellular structures,ultimately leading to aberrant anther development(Li et al.,2004).展开更多
1.Background Crop diseases diminish potential yields by over 20%annually worldwide[1],driving a century-long quest to understand and harness genetic resistance in plants.The genetic basis of disease resistance was fir...1.Background Crop diseases diminish potential yields by over 20%annually worldwide[1],driving a century-long quest to understand and harness genetic resistance in plants.The genetic basis of disease resistance was first recognized in the early 1900s,when Biffen(1905)demonstrated that resistance to stripe rust in wheat(Triticum aestivum)followed Mendelian rules of inheritance[2].This foundational insight was later expanded by Flor(1942),who formulated the“gene-for-gene”model describing the genetic interplay between host resistance(R)genes and pathogen avirulence(Avr)genes[3].展开更多
基金supported by Princess Nourah bint Abdulrahman University,Riyadh,Saudi Arabia through the Researchers Supporting Project PNURSP2025R333.
文摘Human activity recognition(HAR)is a method to predict human activities from sensor signals using machine learning(ML)techniques.HAR systems have several applications in various domains,including medicine,surveillance,behavioral monitoring,and posture analysis.Extraction of suitable information from sensor data is an important part of the HAR process to recognize activities accurately.Several research studies on HAR have utilizedMel frequency cepstral coefficients(MFCCs)because of their effectiveness in capturing the periodic pattern of sensor signals.However,existing MFCC-based approaches often fail to capture sufficient temporal variability,which limits their ability to distinguish between complex or imbalanced activity classes robustly.To address this gap,this study proposes a feature fusion strategy that merges time-based and MFCC features(MFCCT)to enhance activity representation.The merged features were fed to a convolutional neural network(CNN)integrated with long shortterm memory(LSTM)—DeepConvLSTM to construct the HAR model.The MFCCT features with DeepConvLSTM achieved better performance as compared to MFCCs and time-based features on PAMAP2,UCI-HAR,and WISDM by obtaining an accuracy of 97%,98%,and 97%,respectively.In addition,DeepConvLSTM outperformed the deep learning(DL)algorithms that have recently been employed in HAR.These results confirm that the proposed hybrid features are not only practical but also generalizable,making them applicable across diverse HAR datasets for accurate activity classification.
基金supported by Princess Nourah bint Abdulrahman University Researchers Supporting Project number(PNURSP2026R909)Princess Nourah bint Abdulrahman University,Riyadh,Saudi Arabia.
文摘The concept of Human Activity Recognition(HAR)is integral to applications based on Internet of Things(IoT)-enabled devices,particularly in healthcare,fitness tracking,and smart environments.The streams of data from wearable sensors are rich in information,yet their high dimensionality and variability pose a significant challenge to proper classification.To address this problem,this paper proposes hybrid architectures that integrate traditional machine learning models with a deep neural network(DNN)to deliver improved performance and enhanced capabilities for HAR tasks.Multi-sensor HAR data were used to systematically test several hybrid models,including:RF+DNN(Random Forest+Deep Neural Network),XGB+DNN(XGBoost+DNN),GB+DNN(Gradient Boosting+DNN),KNN+DNN(K-Nearest Neighbors+DNN),and DT+DNN(Decision Tree+DNN).The RF+DNN model was the most accurate,achieving a 97.03%score with excellent precision,recall,and F1-score.These findings demonstrate that hybrid machine learning and deep learning systems have a promising future in IoT-based HAR applications.The model provides a novel solution for developing smart and trustworthy monitoring systems that support real-time analytics,patient surveillance,and other IoT applications.
基金supported by the National Key R&D Program of China(No.2023YFC3012002)the National Natural Science Foundation of China(42374073)。
文摘The Bayan Har block,one of China's most seismically active regions,has experienced multiple major earthquakes(≥M 7.0)in recent years.It is a key area for investigating the interactions between the Qinghai-Xizang(Qingzang)Plateau and adjacent blocks,plateau uplift,and strong earthquake mechanisms.P-wave velocity and crustal composition provide key constraints on the properties of distinct tectonic units and their evolutionary modification processes.Based on the results of 8 Deep Seismic Sounding(DSS)profiles completed in the Bayan Har block and surrounding areas over the past 20 years,We constructed one-dimensional P-wave velocity models for the crust of Bayan Har block,Qilian fold belt,Qinling fold belt,Alxa block,Ordos block and Sichuan basin.Furthermore,crustal composition models for different tectonic units were established based on these results.The results reveal that the crustal thickness of the Bayan Har block gradually decreases towards the NNE,NE,and SE directions,while the average crustal velocity increases correspondingly.The felsic layer in the crust accounts for more than half of the total crustal thickness.The mafic content within the crust of different tectonic units exhibits notable variations,which may reflect that the Bayan Har block,Qilian fold belt,and Qinling fold belt have experienced more intensive lithospheric evolution processes compared to Ordos basin and Sichuan basin.The seismicity distribution in this region is significantly controlled by crustal velocity and composition heterogeneity across the Bayan Har block and adjacent areas,which demonstrates that earthquakes within and around the Bayan Har block exhibit both high frequency and larger magnitudes.These seismic characteristics primarily result from intense crustal stress accumulation and release during the outward expansion of the Qingzang Plateau.
文摘The ISO Annual Meeting is the world’s premier event for the international standards community.This year’s event is hosted by RSB,the national standards body of Rwanda.The theme“United for impact”calls us to harness our collective strength in a world that needs bold,lasting change.The ISO Annual Meeting convenes global leaders and change-makers to explore how international standards can unlock progress,foster trust and drive lasting solutions to our shared challenges.
基金financed by“Conselho Nacional de Desenvolvimento Cientifico e Tecnologico”(CNPq)and“Coordenacao de Aperfeicoamento de Pessoal de Nivel Superior”(CAPES)-Finance Code 001.
文摘The increase in wood and wood-based products in the construction and furniture sectors has grown exponentially,generating severe environmental and socioeconomic impacts.Particleboard panels have been the main cost-benefit option on the market due to their lightness and lower cost compared to solid wood.However,the synthetic adhesives used in producing traditional particleboard panels cause serious harm to human health.Developing particleboard panels with fibrous waste and natural adhesives could be a sustainable alternative for these sectors.The work aimed to create particleboards with fibrous wastes from the pseudostem of the banana tree(Musa paradisiaca)and different proportions of the natural adhesive cassava starch-CS in replacement of synthetic adhesive urea-formaldehyde-UF.Five experimental groups were manufactured with banana trees and different percentages of UF and CS adhesives,namely(100UF–0%CS),(50%UF–50%CS),(30%UF–70%CS),(10%UF–90%CS)and(0%UF–100%CS).The particleboards had their physical-mechanical properties determined.The apparent density values did not show significant variation between the assessed treatments.Regarding the water absorption and thickness swelling,the best performances were observed for the panels made without the addition of CS(100%UF).For the mechanical properties of static bending strength and Janka hardness,it was identified that adding up to 50%CS did not interfere with the quality of the panels.These analyses show that the particleboard panels produced with wastes of the banana tree bonded with natural CS adhesivemay be an economically viable and environmentally correct alternative,positively strengthening the development of sustainable strategies.
文摘Human Activity Recognition(HAR)represents a rapidly advancing research domain,propelled by continuous developments in sensor technologies and the Internet of Things(IoT).Deep learning has become the dominant paradigm in sensor-based HAR systems,offering significant advantages over traditional machine learning methods by eliminating manual feature extraction,enhancing recognition accuracy for complex activities,and enabling the exploitation of unlabeled data through generative models.This paper provides a comprehensive review of recent advancements and emerging trends in deep learning models developed for sensor-based human activity recognition(HAR)systems.We begin with an overview of fundamental HAR concepts in sensor-driven contexts,followed by a systematic categorization and summary of existing research.Our survey encompasses a wide range of deep learning approaches,including Multi-Layer Perceptrons(MLP),Convolutional Neural Networks(CNN),Recurrent Neural Networks(RNN),Long Short-Term Memory networks(LSTM),Gated Recurrent Units(GRU),Transformers,Deep Belief Networks(DBN),and hybrid architectures.A comparative evaluation of these models is provided,highlighting their performance,architectural complexity,and contributions to the field.Beyond Centralized deep learning models,we examine the role of Federated Learning(FL)in HAR,highlighting current applications and research directions.Finally,we discuss the growing importance of Explainable Artificial Intelligence(XAI)in sensor-based HAR,reviewing recent studies that integrate interpretability methods to enhance transparency and trustworthiness in deep learning-based HAR systems.
基金supported by the Project of National Key Research and Development Program‘Strategic Science and Technology Innovation Cooperation’Key Special Project(Grant No.2023YFE0206900)the 2115 Talent Development Program of China Agricultural University,the 111 project(B17043)the Project of Yazhouwan Scientific,Technological Administration of Sanya.
文摘Male sterile lines serve as a pivotal tool in plant breeding,offering an effective strategy for developing hybrid cultivars.In cucumber(Cucumis sativus L.),the creation of male sterile lines,most of which exhibit abnormal pollen development,is essential for harnessing heterosis.Accumulating evidence suggests that male sterility in plants is linked to the overaccumulation of Reactive Oxygen Species(ROS),which causes oxidative damage to biomolecules and cellular structures,ultimately leading to aberrant anther development(Li et al.,2004).
基金supported by the National Natural Science Foundation of China (U21A20224)the Key Research and Development Program of Ministry of Science and Technology of China(2023YFD1200402) to Zhiyong Liuby funding from King Abdullah University of Science and Technology to Brande B.H.Wulff
文摘1.Background Crop diseases diminish potential yields by over 20%annually worldwide[1],driving a century-long quest to understand and harness genetic resistance in plants.The genetic basis of disease resistance was first recognized in the early 1900s,when Biffen(1905)demonstrated that resistance to stripe rust in wheat(Triticum aestivum)followed Mendelian rules of inheritance[2].This foundational insight was later expanded by Flor(1942),who formulated the“gene-for-gene”model describing the genetic interplay between host resistance(R)genes and pathogen avirulence(Avr)genes[3].
