Presidency University

Object detection serves as a fundamental task in the field of computer vision and recent developments in YOLO family aim to enhance the real time detection performance. However, the generalization performance of recent YOLO models on unseen video datasets remains underexplored. Analysing the model performance on unseen datasets is essential for assessing robustness in real world deployments. This work provides a systematic comparison of recent YOLO versions and their variants. The study evaluates YOLOv9, YOLOv10 and YOLOv11 models on an unseen MOT20 video dataset for multi pedestrian detection. Pedestrian detection is important because it forms the basis for many computer vision tasks involving human interaction, crowd monitoring, behaviour analysis, and traffic management the models and their variants are evaluated using the metrics: recall, inference speed and GFLOPs. The experimental results indicates that the variant YOLOv9-m achieves highest recall of 43.9% among all evaluated models, while YOLOv11-n showed marginally lower recall value of 40.9%. However, YOLOv11-n exhibits significantly faster inference speed (9.6ms per image) compared to YOLOv9-m (26.3ms per image) and fewer computational resources-(6.5 Vs 131.3). In contrast YOLOv10 exhibits significantly lower recall (28%) despite its increased efficiency. These findings highlight the inherent trade-offs between accuracy-efficiency in recent YOLO architectures. The study offers the understanding of strengths and limitations of modern YOLO models, aiding in model selection for real time computer vision applications.

Side-Channel Attacks (SCA) represent a critical and increasingly significant class of security vulnerabilities in modern computing and electronic systems, exploiting indirect information leakage rather than traditional algorithmic weaknesses. Unlike conventional cryptanalytic attacks, which target the mathematical structure of cryptographic algorithms, SCAs leverage unintentional physical or logical emissions produced by devices during computation. These emissions may include timing information, power consumption patterns, electromagnetic radiation, acoustic signals, and even subtle variations in system temperature. The underlying principle is that any physical instantiation of a computational process inherently produces observable side effects correlated with internal data, including secret keys or sensitive algorithmic states. By monitoring these side effects, attackers can infer confidential information, effectively bypassing the theoretical security guarantees of cryptographic primitives.

Environmental systems are inherently complex and uncertain, influenced by numerous interacting factors including weather variability, pollutant emissions, and human activities. Traditional deterministic modelling approaches often fail to capture both the random variability of these systems and the imprecision inherent in expert knowledge. This paper presents a comprehensive review and application of hybrid fuzzy–stochastic modelling techniques for environmental risk assessment and pollution prediction. Fuzzy logic provides a framework to incorporate qualitative and linguistic information, allowing expert judgments and regulatory standards to be expressed in interpretable terms such as “high risk” or “moderate contamination.” Stochastic mathematics captures random variations in environmental variables, representing uncertainty through probability distributions, simulations, and scenario analysis. The integration of these approaches enables dual representation of uncertainty, combining probabilistic risk assessment with qualitative reasoning, which enhances predictive accuracy and supports informed decision-making. The paper discusses practical applications across air and water pollution management, soil contamination, and climate-related risk assessment, highlighting the value of hybrid models in situations with incomplete data or high variability. Advancements in adaptive fuzzy rule-based systems, high-resolution stochastic simulations, and data-driven calibration techniques are examined, demonstrating how these innovations improve the responsiveness, flexibility, and interpretability of environmental models. Case studies illustrate the effectiveness of hybrid models in predicting urban air quality, contaminant transport in water bodies, and the impact of extreme environmental events. Finally, the paper identifies future research directions, including integration with real-time sensor data, multi-scale modelling, artificial intelligence optimisation, and enhanced visualisation techniques. Overall, the hybrid fuzzy–stochastic framework offers a robust and versatile tool for sustainable environmental management, providing decision-makers with comprehensive, interpretable, and actionable insights into complex environmental risks. The approach facilitates both immediate operational decisions and long-term planning, supporting the development of resilient, adaptive, and scientifically informed environmental policies.

This is an empirical study of 3 government schools in the Srikakulam District of Andhra Pradesh- ZPHS Ippili, A.P.M.S. Tamada, and Government High School Ranasthalam. It attempts to examine the causal elements behind the school drop out rates in this rural coastal region of Andhra Pradesh. Regardless of the introduction of "RTE Act, 2009 " and other government programs like "Sarva Shiksha Abhiyan, Mid-day Meal Scheme, and National Educational Policy 2020" dropout rates are still above the national average. This research uses mixed research design which incorporates qualitative and quantitative data. The results reveal that the causes of high rates of dropout are mostly domestic obligations, inadequate infrastructure, parental ignorance, and poverty especially in fishing communities. The role of the Panchayat Education committees, the uneven execution of the educational programmes and the lack of awareness regarding the same initiatives is also noted in this research. It supports the intensification of support programs, through strengthening of support programs like the Talliki Vandanam, by improving school infrastructure, through empowering the parents, via educational outreach, and encouraging local non-governmental organizations to work together with government bodies. The study finds that the solution to the issue of school dropout is a multi-faceted approach that the various stakeholders in the education industry should be the family, school, and local self-governmental bodies. Keywords: Government Schemes, Panchayati Raj, Parental Negligence, Rural Education, School Dropouts