Faculty Publications
Permanent URI for this communityhttp://192.168.24.11:4000/handle/123456789/7
Scholarly Publications by Integral Academia
Browse
3 results
Search Results
Item Study of an Innovative Approach to IoT Based Human Activity Recognition(BP International, 2025) Motashim Rasool , Rizwan Akhtar , Uvais AhmadRecognizing 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.Item A Comprehensive Analysis of Vulnerabilities and Cybersecurity Issues for Digital Agriculture and Food Industry(Taylor & Francis CRC Press, 2025) Gausiya Yasmeen, Syed Adnan Afaq, Saman UzmaThe foundation of practically all countries in the globe is agriculture. Every cycle's area, from jobs to satiating food needs, is affected. The world's population is expanding more swiftly, which necessitates feeding more stomachs. To fulfill the growing demand for food, smart technology like IoT – Internet of Things – is already being applied in agriculture. Due to the widespread use of IoT technologies, virtually every sector of the economy, including food and agriculture, is also affected by digital transformation. Smart gadgets are routinely utilized by an extensive throng, including farmers and business owners. Such kinds of instruments are utilized in varied ways, from employing drones to assist with tasks like pesticide spraying to continuously nursing the health of harvests and the humidity content of the soil. The advantages of smart agriculture are immense; the condition of soil and seed traits can be observed using smart data, and new technology can provide conveniences never before possible while also enhancing the management and standard of agricultural farming. The advancement of agriculture to smart farming has resulted in an increase in cyber threats. Smart farming ecosystems are, however, exposed to a variety of cybersecurity threats and vulnerabilities due to the employing of IoT and smart communication technologies. Ransomware is the cyberattack that affects the food industry the most frequently right now. Such hacks may cause economic instability in countries that heavily rely on agriculture. As per reports food inflation has risen up to 4.5% which was earlier 3% in the year 2023. Moreover, this year was declared “El Nino” which can result in growth of the hunger index in poor nations. So, modern cultures' food supply chains are essential components. It employs network connections and is fully digitalized, just like other facets of life. Nevertheless, the widespread practice of technical know-how comes with built-in security dangers and weaknesses, and both industries are now under more scrutiny than ever. The chapter presents a complete study on the protection and confidentiality related to food industries and agriculture sector data.Item Futuristic Trends In Electronics and Instrumentation Engineering(Iterative International Publishers (IIP), Selfypage Developers Pvt Ltd., 2025) Sathya P., Ranjan Maheshwari, Ranjeeth M., Archana YadavThis book series aims to bring together researchers and practitioners from academia and industry to focus on recent systems and techniques in the broad field of electronics, instrumentation and communication Engineering. Original research papers, state-of-the-art reviews are included in all areas of Electronics & Instrumentation Engineering. It also focuses on a range of issues on Semiconductor devices, non-conventional energy resources, analog and Digital Circuits, Optical Networks, and communication, etc.