Communities in DSpace
Select a community to browse its collections.
Recent Submissions
Study of an Innovative Approach to IoT Based Human Activity Recognition
(BP International, 2025) Motashim Rasool , Rizwan Akhtar , Uvais Ahmad
Recognizing human activities is vital for numerous contemporary applications rooted in the Internet of Things (IoT) framework, spanning from the creation of intelligent video surveillance setups to the advancement of robotic assistants for the elderly. Recently, there has been significant exploration into machine learning algorithms to enhance the recognition of human activities. Despite these research endeavors, there remains a notable dearth of studies focusing on efficiently recognizing complex human activities, particularly those involving transitions, and no research has been conducted to assess the impact of noise in training data on algorithm performance. This paper addresses these gaps by presenting an innovative activity recognition system centered on a neural classifier with memory capabilities, designed to optimize the classification of both transitional and non-transitional human activities. Utilizing unobtrusive IoT devices such as accelerometers and gyroscopes integrated into widely-used smartphones, the system effectively identifies human activities [1,2]. The key feature of the proposed system lies in leveraging a neural network augmented with short-term memory to retain information about preceding activities' characteristics. Experimental validation demonstrates the reliability and accuracy of the proposed system compared to state-of-the-art classifiers, highlighting its robustness in handling noisy data. Human Activity Recognition (HAR) is essential for various modern applications within the Internet of Things (IoT) framework, from developing intelligent video surveillance systems to enhancing robotic assistants for the elderly. Despite significant advancements in machine learning algorithms for HAR, there is a notable lack of research on effectively recognizing complex human activities, particularly those involving transitions, and assessing the impact of noise in training data on algorithm performance. This paper addresses these gaps by presenting an innovative activity recognition system centered on a neural classifier with memory capabilities. Designed to optimize the classification of both transitional and non-transitional human activities, the system employs unobtrusive IoT devices such as accelerometers and gyroscopes integrated into widely used smartphones [1,3]. A key feature of the proposed system is the utilization of a neural network augmented with short-term memory to retain information about preceding activities' characteristics. Experimental validation demonstrates the system's reliability and accuracy compared to state-of-the-art classifiers, emphasizing its robustness in handling noisy data.
Smart Healthcare Systems Using Computer Vision and IoE
(IGI Global, 2025) Shweta Dwivedi, Farooq Ahamad, Soumya Singh, Syed Adnan Afaq, Vishal Agarwal
Integrating Computer Vision and the Internet of Everything (IoE) in smart healthcare systems represents a transformative shift towards more personalized, efficient, and proactive medical care. Computer Vision technologies enable the analysis and interpretation of visual data, such as medical imaging and patient monitoring, to support diagnostics, treatment planning, and patient management. Meanwhile, IoE extends the capabilities of IoT by incorporating people, processes, and data into a unified ecosystem, enhancing connectivity and data sharing among healthcare devices and systems. This abstract explores the synergies between Computer Vision and IoE in advancing smart healthcare solutions, focusing on how these technologies collectively improve healthcare delivery through enhanced monitoring, diagnostics, and personalized treatment. The review highlights key advancements, applications, and challenges in this field, providing insights into future research directions and the potential for further innovation in smart healthcare systems.
Sustainable planet: Leveraging artificial intelligence for environmental conservation and social well-being
(CRC Press, 2025) Sustainable planet Leveraging artificial intelligence for environmental conservation and social well-being
Sustainability has become more important today than ever before as the world goes through unprecedented environmental challenges. The climate crisis has reached a critical phase and needs immediate action rather than further debate and discussion. AI refers to the abilities of machines to perform just like humans, including learning capabilities by example, recognizing objects, understanding and generating languages, making decisions, and even solving problems amongst others etc., Thus, the integration of artificial intelligence (AI) across sectors has emerged as a promising solution to address critical issues related to climate change, waste, and the environment the destruction of the source. This paper presents a comprehensive analysis of APs contribution to a sustainable planet. AI’s incredible contribution to various industries shows that it can transform our planet towards sustainability. However, successful implementation requires collaborative efforts from governments, industries, academia, and society at large. Embracing APs potential responsibly can lead to a greater sustainable, resilient, and equitable international for contemporary and future generations.
A comprehensive overview of digital image processing techniques in precision agriculture
(CRC Press, 2026) Gausiya Yasmeen, Tasneem Ahmed, Nayyar Ali Usmani
India, a country with an agriculture-based economy, is embracing Precision Agriculture (PA) techniques to optimize the use of seeds, water, and energy. PA uses tools like positioning technology, geographical information systems, satellite navigation, and digital image processing to provide food security and sustainable development. Proper crop monitoring is essential for improving production quality and quantity. PA provides reliable real-time and concurrent information, aiding decision-making. Previously, manual crop monitoring was inefficient due to its time-consuming nature. Digital image processing techniques can be used instead. This article reviews digital image processing-based crop monitoring using satellite-driven satellites and Remotely Piloted Aerial Systems (RPAS)-drone-based methods. RPAS-based images yield more effective crop monitoring results than space station images. Digital image processing is used in agricultural applications for controlling undesirable plants, plant diseases, nutrients, pets, and other issues. Future studies using digital image processing in photogrammetry, vegetation indices, crop monitoring applications, and a variety of crops have opened opportunities for further research in this field.
Leveraging machine learning for emerging trends in Information Technology: A review
(CRC Press, 2025) Shweta Dwivedi, Syed Adnan Afaq, Saurabh Srivastava, Uma Gupta Garg, Saman Uzma
In the continually changing information technology (IT) field, staying on top of emerging trends is critical for firms looking to preserve a competitive advantage. Machine Learning (ML) approaches have evolved as practical tools for evaluating large volumes of data and generating valuable insights. This review paper examines the convergence between machine learning (ML) and new IT trends, emphasizing cyber security, edge computing, the Internet of Things (IoT), cloud computing, and quantum computing. This study examines recent research and case examples to demonstrate how ML algorithms are being used to address difficulties and create opportunities in specific sectors. Furthermore, it explores the ramifications for corporations, academia, and governments, offering insights into future trends and prospective study areas.