The rapid advancement of image manipulation techniques, including copy–move, splicing, and AIgenerated forgeries, has intensified the need for reliable image forgery detection and localization methods. Recent deep learning approaches have significantly improved performance through convolutional encoder–decoder architectures, multi-scale feature fusion models, transformer-based global attention mechanisms, contrastive learning frameworks, and edge-guided localization strategies. Although these methods achieve strong results on benchmark datasets such as CASIA and CoMoFoD, challenges remain in cross-dataset generalization, interpretability, and robustness to diverse manipulation types. Furthermore, heavy reliance on pixel-level supervision limits scalability in practical scenarios. This review systematically analyzes state-of-the-art techniques from 2023 to 2025, categorizing them based on architectural design, supervision strategy, and feature modeling mechanisms. A comparative evaluation of benchmark datasets and performance metrics is presented, highlighting emerging trends toward hybrid CNN–Transformer architectures, multiscale contextual modeling, and weakly supervised frameworks. The paper concludes by identifying open research challenges and future directions for developing more robust and generalizable forgery localization systems.
Market awareness is a critical determinant of organizational competitiveness in dynamic business environments. This study examines the level of market awareness among employees of Mithra Agencies Pvt. Ltd., a leading Maruti Suzuki dealership in Hyderabad. Using a descriptive research design, primary data were collected from 100 employees through structured questionnaires, supplemented by secondary data from company records and published literature. The study analyses awareness of products, market trends, competitor activities, customer preferences, and internal communication effectiveness. Findings reveal moderate to high awareness levels but identify gaps in training frequency, competitor analysis, and access to market intelligence. The paper concludes that strengthening training programs, communication systems, and market research practices can significantly enhance organizational responsiveness and performance.
Mental health disorders such as stress, anxiety, and mental fatigue have become increasingly prevalent in modern society due to rapid technological advancements, increased workload, and prolonged screen exposure. Continuous stress without proper monitoring can lead to severe physical and psychological health problems. Traditional stress detection methods rely on manual observation, questionnaires, or clinical diagnosis, which are often timeconsuming and unsuitable for real-time monitoring. This paper proposes a real-time mental health monitoring system that utilizes multimodal artificial intelligence techniques combined with webcam-based facial analysis and Internet of Things (IoT) physiological sensing. The system collects facial expressions through a webcam and physiological parameters such as heart rate and body temperature through IoT sensors. Machine learning and deep learning algorithms process the collected data to identify stress patterns and classify stress levels accurately. The system generates alerts when abnormal stress levels are detected and stores data for further analysis. The proposed system improves detection accuracy by integrating multiple data sources and enables continuous monitoring of mental health conditions. The system can be implemented in workplaces, educational institutions, healthcare facilities, and remote monitoring environments to support early stress detection and preventive healthcare.
Transformer bushings and air-end insulators are essential components in high-voltage electrical power systems, providing both electrical insulation and mechanical support for conductors passing through grounded equipment. Conventional bushings are generally manufactured using porcelain ceramic materials due to their high dielectric strength and good mechanical properties. However, porcelain insulators possess inherent limitations such as high weight and brittle fracture behavior, which make them susceptible to mechanical failure under impact loading, environmental stress, and operational disturbances. These failures can lead to equipment damage, power outages, and increased maintenance costs in substations. This study focuses on the analysis and optimization of transformer bushings to improve their impact resistance and structural reliability. The mechanical behavior of conventional porcelain bushings is investigated under different loading conditions including internal pressure, bending load, and impact forces. A mathematical analysis is performed to evaluate the stress distribution and structural response of the bushing structure. Furthermore, the potential replacement of porcelain material with polymer-based composite materials such as glass fiber reinforced polymers is explored to enhance impact resistance while reducing overall weight. Finite Element Analysis (FEA) is proposed to simulate stress, deformation, and failure behavior under various operating conditions. The results of this study are expected to contribute to the development of lightweight, durable, and impact-resistant transformer bushings suitable for modern high-voltage power systems.
This paper presents the design, hardware implementation, and experimental validation of a low-cost Automatic Power Factor Controller (APFC) integrated with Internet of Things (IoT)-based real-time remote monitoring. The system employs zero-crossing detection (ZCD) circuits for both supply voltage and load current to compute the displacement power factor continuously on an Arduino Uno (ATmega328P) microcontroller. Based on the measured phase lag between voltage and current zero crossings, the controller automatically switches up to two relaycontrolled capacitor banks to compensate reactive power and maintain power factor at or above 0.95 lagging. An ESP32 Wi-Fi module receives electrical parameter data from the Arduino over UART and serves a responsive HTML5 web dashboard accessible from any browser on the local network without internet dependency. The system also features a dual-mode operation — fully automatic capacitor switching and manual override via push buttons — with a 16×2 LCD for on-panel display. Experimental results confirm power factor improvement from 0.823 to 0.994 under mixed inductive-resistive loading at 230 V, 50 Hz, representing a 78.0% reduction in reactive power demand. The complete prototype cost is approximately INR 2,500, offering a practical and scalable solution for small industries and educational institutions.
