159 research outputs found
DeepIOD: Towards A Context-Aware Indoor–Outdoor Detection Framework Using Smartphone Sensors
No full textAccurate indoor–outdoor detection (IOD) is essential for location-based services, context-aware computing, and mobile applications, as it enhances service relevance and precision. However, traditional IOD methods, which rely only on GPS data, often fail in indoor environments due to signal obstructions, while IMU data are unreliable on unseen data in real-time applications due to reduced generalizability. This study addresses this research gap by introducing the DeepIOD framework, which leverages IMU sensor data, GPS, and light information to accurately classify environments as indoor or outdoor. The framework preprocesses input data and employs multiple deep neural network models, combining outputs using an adaptive majority voting mechanism to ensure robust and reliable predictions. Experimental results evaluated on six unseen environments using a smartphone demonstrate that DeepIOD achieves significantly higher accuracy than methods using only IMU sensors. Our DeepIOD system achieves a remarkable accuracy rate of 98–99% with a transition time of less than 10 ms. This research concludes that DeepIOD offers a robust and reliable solution for indoor–outdoor classification with high generalizability, highlighting the importance of integrating diverse data sources to improve location-based services and other applications requiring precise environmental context awareness
Hydraulic simulations to evaluate and predict design and operation of the Chashma Right Bank Canal
No full textIrrigation systems / Irrigation canals / Flow control / Velocity / Canal regulation techniques / Hydraulics / Simulation models / Design / Operations / Crop-based irrigation / Distributary canals / Water delivery / Policy / Protective irrigation / Water allocation / Water requirements / Sedimentation / Water distribution / Equity / Water conveyance / Pakistan / Chashma Right Bank Canal
DeepILS: Toward Accurate Domain-Invariant AIoT-Enabled Inertial Localization System
No full textAccurate indoor localization and navigation enable real-time, ubiquitous, location-based services. Over the past decade, data-driven approaches for inertial odometry have shown the potential to enhance indoor positioning accuracy. However, low-cost inertial measurement units (IMUs), commonly used in smartphones and IoT devices, are prone to significant noise, leading to drift and degraded performance in navigation algorithms. This article presents a novel, lightweight, and real-time end-to-end framework, DeepILS Brossard et al., (2020), designed to process raw inertial data for precise pedestrian localization in indoor environments. DeepILS utilizes a residual network enhanced with channel-wise and spatial attention mechanisms, enabling accurate velocity and position estimation across diverse motion dynamics. The framework's effectiveness is validated using four benchmarks and two newly introduced datasets in real-time edge scenarios. These datasets were collected across diverse indoor environments at the KAIST campus and Incheon National Airport, using multiple hardware platforms, including the KAIST IoT positioning module and Android smartphones. Experimental results, including tests on unseen data and comprehensive ablation studies, demonstrate that DeepILS improves localization accuracy by 70% compared to state-of-the-art methods while effectively mitigating sensor noise and enhancing robustness in real-world environments. Specifically, DeepILS exhibits excellent edge performance on IoT devices, making it highly suitable for real-time applications.
Child Pedestrian Safety: Study of Street-Crossing Behaviour of Primary School Children with Adult Supervision
No full textRoad traffic accidents are the primary cause of injuries and fatalities among children. The current study focuses on children’s (un)safe crossing behaviour in a real traffic situation accompanied by an adult at a crosswalk in front of their school. The study aims to investigate if there are differences in crossing behaviour related to road infrastructure (i.e., one-way and two-way street, elevated and non-elevated street crossing), the gender of the child, and the effect of the accompanying adult’s behaviour on the child’s crossing behaviour. Primary school children from two urban schools in Flanders (Belgium) were observed for three days while crossing the street in front of their school in the morning and afternoon. A total of 241 child–adult pairs were observed. Descriptive analysis, Pearson chi-square tests, and binary logistic regression models were used to find differences between groups. More than half of the crossings exhibited two or more unsafe behaviours. Not stopping at the curb before crossing was the most unsafe behaviour, exhibited by 47.7% of children; not looking for oncoming traffic before and during the crossing was the second most unsafe behaviour, exhibited by 39.4% of the children. The only difference between boys’ and girls’ crossing behaviour was in stopping at the curb with girls 1.901 times more likely to stop before crossing as compared to boys. Adults holding hands of the child resulted in safer behaviours by children. The children not holding hands displayed significantly riskier behaviour in running or hopping while crossing the street and being distracted. The study reinforces the need to improve the transportation system through infrastructural interventions (elevated crosswalks), as well as educating and training children and the parents on safe crossing behaviour in traffic
Assessment of phenotypic diversity in the USDA collection of quinoa links genotypic adaptation to germplasm origin
No full textQuinoa’s germplasm evaluation is the first step towards determining its suitability under new environmental conditions. The aim of this study was to introduce suitable germplasm to the lowland areas of the Faisalabad Plain that could then be used to introduce quinoa more effectively to that region. A set of 117 quinoa genotypes belonging to the USDA quinoa collection was evaluated for 11 phenotypic quantitative traits (grain yield (Y), its biological and numerical components plus phenological variables) in a RCBD during two consecutive growing seasons at the University of Agriculture, Faisalabad, Pakistan under mid-autumn sowings. Genotypic performance changed across the years, however most phenotypic traits showed high heritability, from 0.75 for Harvest Index (HI) to 0.97 for aerial biomass (B) and Y. Ordination and cluster analyses differentiated four Academic Editors: Cataldo Pulvento and Didier Bazile Received: 28 January 2022 Accepted: 3 March 2022 Published: 10 March 2022 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Copyright: © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). groups dominated by genotypes from: Peru and the Bolivian Highlands (G1); the Bolivian Highlands (G2); the Ballón collection (regarded as a cross between Bolivian and Sea Level (Chilean) genotypes) plus Bolivian Highlands (G3); and Ballón plus Sea Level (G4), this latter group being the most differentiated one. This genetic structure shared similarities with previous groups identified using SSR markers and G×Edata from an international quinoa test. G4 genotypes showed the highest Y associated with higher B and seed numbers (SN), while HI made a significant contribution to yield determination in G2 and seed weight (SW) in G3. G1 and G2 showed the lowest Y associated with a lower B and SN. Moreover, SW showed a strongly negative association with SN in G2. Accordingly, G4followed by G3 are better suited to the lowland areas of Faisalabad plain and the physiological traits underlying yield determination among genotypic groups should be considered in future breeding programs.Fil: Hafeez, Muhammad Bilal. University Of Agriculture; PakistánFil: Iqbal, Shahid. University Of Agriculture; PakistánFil: Li, Yuanyuan. Shandong Normal University. College of Life Science. Shandong Provincial Key Laboratory of Plant Stress Research; ChinaFil: Saddiq, Muhammad Sohail. Ghazi University. Department of Agronomy; PakistánFil: Basra, Shahzad M. A.. University Of Agriculture; PakistánFil: Zhang, Hui. Shandong Normal University. College of Life Science. Shandong Provincial Key Laboratory of Plant Stress Research; ChinaFil: Zahra, Noreen. University Of Agriculture; PakistánFil: Akram, Muhammad Z.. University Of Agriculture; PakistánFil: Bertero, Hector Daniel. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina. Universidad de Buenos Aires. Facultad de Agronomía. Departamento de Producción Vegetal. Cátedra de Producción Vegetal; ArgentinaFil: Curti, Ramiro Nestor. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Salta; Argentina. Universidad Nacional de Salta. Facultad de Ciencias Naturales. Escuela de Agronomía. Laboratorio de Investigaciones Botánicas; Argentin
실내외 감지를 위한 대조 학습 표현을 사용한 표 형식 데이터 분류
No full text학위논문(박사) - 한국과학기술원 : 전산학부, 2025.2,[vi, 83 p. :]The primary goal of indoor-outdoor detection (IOD) is to accurately determine whether a user is indoors or outdoors using smartphone sensors, crucial for applications like pedestrian navigation, activity recognition, and Internet of Things (IoT) device management. Existing methods struggle with scalability and effectiveness because they rely on large, labeled datasets that are costly and difficult to obtain, particularly in dynamic environments, and often fail to capture temporal and spatial dependencies in unseen conditions. To address these issues, we propose TabCLR, a novel self-supervised learning (SSL) framework designed for IOD. TabCLR leverages contrastive learning to utilize unlabeled data effectively, introducing several innovations over existing methods like SCARF, including Random Permutation of Features and Sampling from Marginal Distributions (RPFC) for realistic feature corruption, a Spatial-Context Contrastive Loss (SCCL) function tailored for spatial relationships in IOD, and the incorporation of self-attention within the encoder to capture temporal dependencies. These advancements allow TabCLR to enhance its ability to generalize across different environments and datasets, even in the absence of labeled data. Evaluated against other IOD works and SCARF, the state-of-the-art SSL approach, on multiple datasets, including the CIOD and newly proposed MIOD, KIOD, and DIOD, TabCLR demonstrates superior performance and robustness, advancing IOD technology, and improving system accuracy and efficiency.한국과학기술원 :전산학부
Collected Papers (Neutrosophics and other topics), Volume XIV
No full textThis fourteenth volume of Collected Papers is an eclectic tome of 87 papers in Neutrosophics and other fields, such as mathematics, fuzzy sets, intuitionistic fuzzy sets, picture fuzzy sets, information fusion, robotics, statistics, or extenics, comprising 936 pages, published between 2008-2022 in different scientific journals or currently in press, by the author alone or in collaboration with the following 99 co-authors (alphabetically ordered) from 26 countries: Ahmed B. Al-Nafee, Adesina Abdul Akeem Agboola, Akbar Rezaei, Shariful Alam, Marina Alonso, Fran Andujar, Toshinori Asai, Assia Bakali, Azmat Hussain, Daniela Baran, Bijan Davvaz, Bilal Hadjadji, Carlos Díaz Bohorquez, Robert N. Boyd, M. Caldas, Cenap Özel, Pankaj Chauhan, Victor Christianto, Salvador Coll, Shyamal Dalapati, Irfan Deli, Balasubramanian Elavarasan, Fahad Alsharari, Yonfei Feng, Daniela Gîfu, Rafael Rojas Gualdrón, Haipeng Wang, Hemant Kumar Gianey, Noel Batista Hernández, Abdel-Nasser Hussein, Ibrahim M. Hezam, Ilanthenral Kandasamy, W.B. Vasantha Kandasamy, Muthusamy Karthika, Nour Eldeen M. Khalifa, Madad Khan, Kifayat Ullah, Valeri Kroumov, Tapan Kumar Roy, Deepesh Kunwar, Le Thi Nhung, Pedro López, Mai Mohamed, Manh Van Vu, Miguel A. Quiroz-Martínez, Marcel Migdalovici, Kritika Mishra, Mohamed Abdel-Basset, Mohamed Talea, Mohammad Hamidi, Mohammed Alshumrani, Mohamed Loey, Muhammad Akram, Muhammad Shabir, Mumtaz Ali, Nassim Abbas, Munazza Naz, Ngan Thi Roan, Nguyen Xuan Thao, Rishwanth Mani Parimala, Ion Pătrașcu, Surapati Pramanik, Quek Shio Gai, Qiang Guo, Rajab Ali Borzooei, Nimitha Rajesh, Jesús Estupiñan Ricardo, Juan Miguel Martínez Rubio, Saeed Mirvakili, Arsham Borumand Saeid, Saeid Jafari, Said Broumi, Ahmed A. Salama, Nirmala Sawan, Gheorghe Săvoiu, Ganeshsree Selvachandran, Seok-Zun Song, Shahzaib Ashraf, Jayant Singh, Rajesh Singh, Son Hoang Le, Tahir Mahmood, Kenta Takaya, Mirela Teodorescu, Ramalingam Udhayakumar, Maikel Y. Leyva Vázquez, V. Venkateswara Rao, Luige Vlădăreanu, Victor Vlădăreanu, Gabriela Vlădeanu, Michael Voskoglou, Yaser Saber, Yong Deng, You He, Youcef Chibani, Young Bae Jun, Wadei F. Al-Omeri, Hongbo Wang, Zayen Azzouz Omar
Collected Papers (Neutrosophics and other topics), Volume XIV
No full textThis fourteenth volume of Collected Papers is an eclectic tome of 87 papers in Neutrosophics and other fields, such as mathematics, fuzzy sets, intuitionistic fuzzy sets, picture fuzzy sets, information fusion, robotics, statistics, or extenics, comprising 936 pages, published between 2008-2022 in different scientific journals or currently in press, by the author alone or in collaboration with the following 99 co-authors (alphabetically ordered) from 26 countries: Ahmed B. Al-Nafee, Adesina Abdul Akeem Agboola, Akbar Rezaei, Shariful Alam, Marina Alonso, Fran Andujar, Toshinori Asai, Assia Bakali, Azmat Hussain, Daniela Baran, Bijan Davvaz, Bilal Hadjadji, Carlos Díaz Bohorquez, Robert N. Boyd, M. Caldas, Cenap Özel, Pankaj Chauhan, Victor Christianto, Salvador Coll, Shyamal Dalapati, Irfan Deli, Balasubramanian Elavarasan, Fahad Alsharari, Yonfei Feng, Daniela Gîfu, Rafael Rojas Gualdrón, Haipeng Wang, Hemant Kumar Gianey, Noel Batista Hernández, Abdel-Nasser Hussein, Ibrahim M. Hezam, Ilanthenral Kandasamy, W.B. Vasantha Kandasamy, Muthusamy Karthika, Nour Eldeen M. Khalifa, Madad Khan, Kifayat Ullah, Valeri Kroumov, Tapan Kumar Roy, Deepesh Kunwar, Le Thi Nhung, Pedro López, Mai Mohamed, Manh Van Vu, Miguel A. Quiroz-Martínez, Marcel Migdalovici, Kritika Mishra, Mohamed Abdel-Basset, Mohamed Talea, Mohammad Hamidi, Mohammed Alshumrani, Mohamed Loey, Muhammad Akram, Muhammad Shabir, Mumtaz Ali, Nassim Abbas, Munazza Naz, Ngan Thi Roan, Nguyen Xuan Thao, Rishwanth Mani Parimala, Ion Pătrașcu, Surapati Pramanik, Quek Shio Gai, Qiang Guo, Rajab Ali Borzooei, Nimitha Rajesh, Jesús Estupiñan Ricardo, Juan Miguel Martínez Rubio, Saeed Mirvakili, Arsham Borumand Saeid, Saeid Jafari, Said Broumi, Ahmed A. Salama, Nirmala Sawan, Gheorghe Săvoiu, Ganeshsree Selvachandran, Seok-Zun Song, Shahzaib Ashraf, Jayant Singh, Rajesh Singh, Son Hoang Le, Tahir Mahmood, Kenta Takaya, Mirela Teodorescu, Ramalingam Udhayakumar, Maikel Y. Leyva Vázquez, V. Venkateswara Rao, Luige Vlădăreanu, Victor Vlădăreanu, Gabriela Vlădeanu, Michael Voskoglou, Yaser Saber, Yong Deng, You He, Youcef Chibani, Young Bae Jun, Wadei F. Al-Omeri, Hongbo Wang, Zayen Azzouz Omar
Mouth and oral disease classification using InceptionResNetV2 method
No full textDigital tools have greatly improved the detection and diagnosis of oral and dental disorders like cancer and gum disease. Lip or oral cavity cancer is more likely to develop in those with potentially malignant oral disorders. A potentially malignant disorder (PMD) and debilitating condition of the oral mucosa, oral submucous fibrosis (OSMF), can have devastating effects on one’s quality of life. Incorporating deep learning into diagnosing conditions affecting the mouth and oral cavity is challenging. Mouth and Oral Diseases Classification using InceptionResNetV2 Method was established in the current study to identify diseases such as gangivostomatitis (Gum), canker sores (CaS), cold sores (CoS), oral lichen planus (OLP), oral thrush (OT), mouth cancer (MC), and oral cancer (OC). The new collection, termed "Mouth and Oral Diseases" (MOD), comprises seven distinct categories of data. Compared to state-of-the-art approaches, the proposed InceptionResNetV2 model’s 99.51% accuracy is significantly higher.© The Author(s) 2023. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.fi=vertaisarvioitu|en=peerReviewed
Collected Papers (on various scientific topics), Volume XIII
No full textThis thirteenth volume of Collected Papers is an eclectic tome of 88 papers in various fields of sciences, such as astronomy, biology, calculus, economics, education and administration, game theory, geometry, graph theory, information fusion, decision making, instantaneous physics, quantum physics, neutrosophic logic and set, non-Euclidean geometry, number theory, paradoxes, philosophy of science, scientific research methods, statistics, and others, structured in 17 chapters (Neutrosophic Theory and Applications; Neutrosophic Algebra; Fuzzy Soft Sets; Neutrosophic Sets; Hypersoft Sets; Neutrosophic Semigroups; Neutrosophic Graphs; Superhypergraphs; Plithogeny; Information Fusion; Statistics; Decision Making; Extenics; Instantaneous Physics; Paradoxism; Mathematica; Miscellanea), comprising 965 pages, published between 2005-2022 in different scientific journals, by the author alone or in collaboration with the following 110 co-authors (alphabetically ordered) from 26 countries: Abduallah Gamal, Sania Afzal, Firoz Ahmad, Muhammad Akram, Sheriful Alam, Ali Hamza, Ali H. M. Al-Obaidi, Madeleine Al-Tahan, Assia Bakali, Atiqe Ur Rahman, Sukanto Bhattacharya, Bilal Hadjadji, Robert N. Boyd, Willem K.M. Brauers, Umit Cali, Youcef Chibani, Victor Christianto, Chunxin Bo, Shyamal Dalapati, Mario Dalcín, Arup Kumar Das, Elham Davneshvar, Bijan Davvaz, Irfan Deli, Muhammet Deveci, Mamouni Dhar, R. Dhavaseelan, Balasubramanian Elavarasan, Sara Farooq, Haipeng Wang, Ugur Halden, Le Hoang Son, Hongnian Yu, Qays Hatem Imran, Mayas Ismail, Saeid Jafari, Jun Ye, Ilanthenral Kandasamy, W.B. Vasantha Kandasamy, Darjan Karabašević, Abdullah Kargın, Vasilios N. Katsikis, Nour Eldeen M. Khalifa, Madad Khan, M. Khoshnevisan, Tapan Kumar Roy, Pinaki Majumdar, Sreepurna Malakar, Masoud Ghods, Minghao Hu, Mingming Chen, Mohamed Abdel-Basset, Mohamed Talea, Mohammad Hamidi, Mohamed Loey, Mihnea Alexandru Moisescu, Muhammad Ihsan, Muhammad Saeed, Muhammad Shabir, Mumtaz Ali, Muzzamal Sitara, Nassim Abbas, Munazza Naz, Giorgio Nordo, Mani Parimala, Ion Pătrașcu, Gabrijela Popović, K. Porselvi, Surapati Pramanik, D. Preethi, Qiang Guo, Riad K. Al-Hamido, Zahra Rostami, Said Broumi, Saima Anis, Muzafer Saračević, Ganeshsree Selvachandran, Selvaraj Ganesan, Shammya Shananda Saha, Marayanagaraj Shanmugapriya, Songtao Shao, Sori Tjandrah Simbolon, Florentin Smarandache, Predrag S. Stanimirović, Dragiša Stanujkić, Raman Sundareswaran, Mehmet Șahin, Ovidiu-Ilie Șandru, Abdulkadir Șengür, Mohamed Talea, Ferhat Taș, Selçuk Topal, Alptekin Ulutaș, Ramalingam Udhayakumar, Yunita Umniyati, J. Vimala, Luige Vlădăreanu, Ştefan Vlăduţescu, Yaman Akbulut, Yanhui Guo, Yong Deng, You He, Young Bae Jun, Wangtao Yuan, Rong Xia, Xiaohong Zhang, Edmundas Kazimieras Zavadskas, Zayen Azzouz Omar, Xiaohong Zhang, Zhirou Ma.
- …
