Authors - Thomas K P, Sherly K K Abstract - Permanent Magnet Synchronous Motors (PMSMs) are commonly utilized in electric vehicle (EV) traction systems because of its high efficiency, power density, and reliability. Conventional field-oriented control (FOC) schemes require accurate rotor position and speed information, typically obtained from mechanical sensors, which increase cost and reduce system reliability. Sensor less control techniques based on observer theory have therefore gained significant attention. Among them, sliding mode observers (SMOs) offer strong robustness against parameter variations and external disturbances but suffer from chattering and noise sensitivity. This paper presents an advanced sensor less FOC strategy for PMSM drives using a super-twisting SMO (ST-SMO) for rotor position sensing and estimation of speed. The proposed approach employs a ST-SMO algorithm to achieve the convergence in finite-time while significantly reducing chattering effects. The observer is integrated into a standard FOC framework and evaluated under EV-relevant operating conditions, including low-speed operation and load transients. Comparative performance discussion demonstrates the suitability and the effectiveness of the proposed method for high-efficiency EV traction.
Authors - N. V. Naik, Raga Madhuri Dhulipudi, Marisetti Sandhya, Jadda Anjan Kumar Abstract - Distributed systems rely on data replication to ensure availability, fault tolerance, and scalability across multiple nodes in modern cloud environments. Replication enables systems to maintain continuity even when individual nodes fail or experience network disruptions. However, replication often introduces synchronization delays between primary and replica nodes, known as replication delay. These delays can cause temporary data inconsistency, stale reads, and increased response latency, degrading application performance and user experience. As infrastructures scale to larger clusters, communication overhead, network latency, and workload variability further amplify replication delays, making efficient synchronization increasingly challenging. Traditional replication mechanisms typically rely on static synchronization intervals or sequential update propagation strategies. These approaches fail to adapt to dynamic network conditions and fluctuating workloads, resulting in inefficient data propagation and delayed consistency across nodes. In large scale systems, such limitations may cause bottlenecks, reduced reliability, and inconsistent states during high workload periods or network congestion. Addressing replication delay is critical for maintaining reliability and consistency in distributed environments. Recent research emphasizes intelligent synchronization mechanisms capable of adapting to changing conditions. Adaptive synchronization strategies that monitor network latency, workload intensity, and node communication patterns offer improvements in replication efficiency. By enabling replication decisions that respond dynamically to system behavior, such approaches reduce synchronization delays and improve data consistency across clusters. Enhanced replication efficiency ultimately strengthens reliability, scalability, and operational performance in modern distributed computing platforms operating under variable workload conditions.
Authors - Shaik Shafi, C Santhoshi Abstract - In the recent past, vehicle theft in India has increasing nearly 2.5 times, with more than 2 lakh vehicles stolen annually. The Delhi NCR region alone accounts for over 30% of reported cases, and in Delhi, a vehicle is reportedly stolen approximately every 14 minutes. These alarming trends highlight the ur-gent need for stronger and smarter vehicle security mechanisms. Traditionally, vehicle anti-theft technologies have relied largely on non-biometric approaches such as GPS–GSM tracking modules. Thus, biometric authentication is an emerging security approach that limits vehicle access to authorized individuals by verifying unique biological traits such as fingerprints, facial features, iris pat-terns, or voice. Although this technology significantly strengthens vehicle security, its widespread deployment still faces certain technical and social constraints. Thus in this paper, an IoT enabled biometric ignition system with security alerts is proposed. The proposed model makes use of an ESP32 micro controller and fingerprint sensor to replace traditional keys. The system operates in two stages: first secure door access and secondly engine ignition authorization. Any unauthorized attempts trigger real-time alerts with GPS location via IoT protocols like MQTT or HTTP. Further, cloud integration enables remote monitoring, data storage, and scalability, making suitable for modern intelligent transport systems. In the same way, the fingerprint-based vehicle starter grants the privilege of starting the vehicle only to the registered users, thus deterring theft and ensuring safety. Over all, biometric vehicle ignition is a dependable, economical, and hassle-free solution to access control as well as theft prevention.
Authors - A.Sree Rama Chandra Murthy, T.Gamya Sri, B.Harshitha, G.Vincent Paul Abstract - Accurate forecasting of drug demand is one of the challenging areas in the healthcare service to reduce waste as well as shortages. Some recent studies focused only on predicting drug use demand for regions and hospitals, missing an overall way to combine these forecasts. In this study, a multilevel machine learning framework is presented that merges regional tender demand predictions with monthly and seasonal order forecasting in hospitals and pharmacies. With historical drug usage, the system captures time-based changes, seasonal demands, and also location specific behaviors . Models for regional tenders predict yearly procurement, but models at hospitals and pharmacies try to tell the need of each month, allowing better resource distribution. The rigorous experimental process showed better estimates and forecasting with less error than just making a single-level prediction. This framework helps to make better purchasing decisions and ensures a stable drug supply across healthcare systems. Health departments, hospital chains, and pharmacy groups can benefit from using a model.
Authors - Seamus Lyons Abstract - Methane (CH4) emission from rice paddies is a significant source of greenhouse gas emissions from agriculture. Currently, most models for methane prediction from rice paddies depend on collecting field data and sending it to a server. In this new paradigm, several privacy concerns arise, model scalability is restricted, and a large number of data points are exposed to the attacker. This paper addresses all privacy con cerns by providing an edge-based solution for modeling methane emis sions from rice paddies that leverages data from edge sensors at respec tive locations, while keeping individual sensor data private. The method employs different machine learning (ML) algorithms, including Linear Regression, Random Forest, XGBoost, and a Feedforward Neural Net work (FNN), implemented using TensorFlow Federated (TFF) in both centralized and federated learning (FL) frameworks. The FL-based FNN achieved an R2 score of 0.91, which was superior to both centralized classical and centralized FL models, especially for highly non-IID client side data distributions in sensor datasets. In summary, this paper extends the current literature on modeling methane emissions from rice paddies and provides a comprehensive evaluation of our proposed FL system ar chitecture, an in-depth discussion of the communication resources re quired for FL implementation, and an examination of the effects of abla tion studies on clients’ data heterogeneity. Therefore, the proposed FL approach is efficient and scalable, enabling safe, privacy-preserving modeling of methane emissions from rice paddies to effectively imple ment Climate Smart Agriculture (CSA) and mitigate global warming while supporting sustainable rice cultivation.
Authors - Gia Nghi Thoi, My An Tran, Tram Thi Tuyet Le, Nhat Van Hoang Nguyen, Long Hong Buu Nguyen, Dien Dinh Abstract - Medical diagnosis using Small Language Models (SLMs) of ten suffers from hallucinations and knowledge inconsistency. While re inforcement learning (RL) from knowledge graph feedback offers a po tential solution, pure reinforcement learning strategies often encounter challenges related to sample inefficiency and poor exploration. To address this, a hybrid training pipeline that combines supervised alignment with structural reinforcement is proposed. The method applies knowledge guided supervised fine-tuning (SFT) with hard negatives to refine deci sion boundaries and employs a bipartite-specific reward model to capture interactions between symptoms and diseases. Experiments on multiple medical datasets, including DXY, GMD, and MED-D, demonstrate that this hybrid approach outperforms pure RL methods. By incorporating knowledge graph (KG) information as a structural regularizer, the model achieves improved accuracy, stronger cross-dataset generalization, and reduced overfitting while maintaining strict adherence to diagnostic out put constraints
Authors - Mustafa Icel, Ochilbek Rakhmanov, Ergul Gunerhan, Muhammad Qasim Abstract - Artificial intelligence driven adaptive learning systems progressively operate as knowledge management platforms by collecting, refining, and using learner knowledge to personalize instruction. However, empirical evidence demonstrating how managed knowledge translates into measurable student achievement remains as a question to answer. This study examines the effective ness of AI driven adaptive learning as a knowledge management system in a high school setting. Using de-identified archival data from 182 students across three academic years, the study explores relationships among AI-managed knowledge mastery, engagement, course performance, and standardized assessment out comes. Learning analytics techniques, including descriptive statistics and Pear son correlation analysis, were employed to examine knowledge–performance re lationships. Predictive modeling using multivariable linear regression and Ran dom Forest classification was performed to assess the extent to which knowledge management indicators predict end-of-course achievement and performance lev els. Results indicate that final knowledge mastery is moderately associated with standardized assessment outcomes and is a stronger predictor of achievement than time-on-task alone. While predictive models demonstrate modest accuracy, findings suggest that AI driven knowledge management supports student achievement when integrated within instructional contexts.
Authors - Akshay Kumar, Reena Satpute, Kumar Gaurav, Sanjit Kumar, Edidiong Akpabio, Sudhir Agarmore Abstract - Recent literature has posed LLMs as nonlinear dynamical systems. LLM safety, in these modern LLMs is about the systematic and critical monitoring of logit based oscillations, hidden state rotations and entropy fluctuations. Many of these important factors are spectral proxies for the generation of imaginary eigenvalues. These imaginary eigenvalues are, in a way, determinants of the latent oscillation energy. Though the system in its original state space is inherently nonlinear, through the Koopman operator, we can linearize the evolution in the lifted space of observables. We design a spectral jailbreak detector that has a Sparsely regularized koopman autoencoder as its backbone. We obtain the koopman operator through this SR-KAE, and also obtain the imaginary component of the eigenvalues of that spectral operator, A new risk score metric is proposed that is used to classify prompts as either jailbreak or safe. This becomes a physics-style stability classifier on prompts. We present several test cases, while we discuss the strengths and limitations of this new system.
Authors - Kamala L, Mohan K G Abstract - This paper presents the error performance of digital commu nication systems operating over α-Beaulieu-Xie (α-BX) and its extreme variant, the α-BXe fading channel. A generalized noise model, additive white generalized Gaussian noise (AWGGN), is adopted to account for various practical scenarios including impulsive and Laplacian environ ments. We derive closed-form average bit error rate (ABER) expressions utilizing the Fox-H function. The mathematical expressions derived are validated through numerical integration for binary phase shift keying (BPSK) and binary frequency shift keying (BFSK) modulation schemes. Our results demonstrate the degradation caused by Laplacian noise and characterize the irreducible error floors inherent in the α-BXe model, providing a robust tool for system designers in complex fading environ ments.
Authors - Akshay Kumar, Deepa Thilak Abstract - Smart city apps are growing quickly, which means that there are more real-time, latency-sensitive, and privacy-critical workloads that are hard for traditional single-cloud computing models to handle. In particular, smart mobility and traffic management systems generate large volumes of geographically distributed data that require efficient processing with minimal delay and high reliability. This project proposes a multi-cloud task scheduling framework that protects privacy and uses federated learning to solve these problems. The suggested system turns real-time smart mobility traffic data into abstract scheduling tasks and sends them to different cloud regions using a lightweight, decision-free task broker. Each cloud region has its own local federated scheduler that uses only data that is available in that region to schedule tasks based on latency and congestion. Federated learning is used to work together to improve scheduling policies by safely combining local model updates without sharing raw data. This keeps data private and meets data sovereignty requirements. The system enables improved scalability, reduced response time, fault tolerance, and avoidance of vendor lock-in compared to centralized scheduling approaches. Using a smart mobility dataset to test the proposed method shows that it works well for scheduling tasks quickly and with privacy in mind in multi-cloud settings.
Authors - Thanh-Phuong Ngo, Van-Thanh Huynh, Thai-Son Nguyen Abstract - This paper presents a novel Reversible Data Hiding (RDH) method for dual images. First, secret data is converted into a binary sequence of equal length and then divided into shorter segments to control the amount of data embedded into each pixel. The embedding process uses two copies of the original image to distribute the data, reducing the impact on each image while maintaining overall image quality. During recovery, the original image is restored by averaging the pixel values at corresponding locations in the two stego images, while the embedded data is recovered through a reverse process. Experimental results on grayscale images demonstrate that the method maintains good image quality, achieving a high Peak Signal-to-Noise Ratio (PSNR) across different embedding levels while ensuring accurate recovery of both the secret data and the original image.
Authors - Yasir Abdullah R, Lakshmana Kumar T, Vijaykumar M, Thirunavukkarasu C, Saravanagukhan P, Hariharasuthan M Abstract - In the recent past, vehicle theft in India has increasing nearly 2.5 times, with more than 2 lakh vehicles stolen annually. The Delhi NCR region alone accounts for over 30% of reported cases, and in Delhi, a vehicle is reportedly stolen approximately every 14 minutes. These alarming trends highlight the ur-gent need for stronger and smarter vehicle security mechanisms. Traditionally, vehicle anti-theft technologies have relied largely on non-biometric approaches such as GPS–GSM tracking modules. Thus, biometric authentication is an emerging security approach that limits vehicle access to authorized individuals by verifying unique biological traits such as fingerprints, facial features, iris pat-terns, or voice. Although this technology significantly strengthens vehicle security, its widespread deployment still faces certain technical and social constraints. Thus in this paper, an IoT enabled biometric ignition system with security alerts is proposed. The proposed model makes use of an ESP32 micro controller and fingerprint sensor to replace traditional keys. The system operates in two stages: first secure door access and secondly engine ignition authorization. Any unauthorized attempts trigger real-time alerts with GPS location via IoT protocols like MQTT or HTTP. Further, cloud integration enables remote monitoring, data storage, and scalability, making suitable for modern intelligent transport systems. In the same way, the fingerprint-based vehicle starter grants the privilege of starting the vehicle only to the registered users, thus deterring theft and ensuring safety. Over all, biometric vehicle ignition is a dependable, economical, and hassle-free solution to access control as well as theft prevention.
Authors - Amol Dhumane, Jitendra Chavan, Arijit Dutta, Priyanka Paygude, Aditi Sharma, Datta Takale, Yashwant Dongre Abstract - Depression is a psychiatric condition that is largely common all over the world and greatly influences the emotional stability, cognitive performance and behavior functioning. Computational techniques that can detect the condition early can help to prevent psychological dangers in the long term and ensure timely treatment of the disease. This paper refers to a complete machine learning architecture of automated depression recognition of textual information based on hybrid feature engineering and ensemble learning approaches. The suggested methodology is a combination of text preprocessing, Term Frequency / Inverse Document Frequency (TF -IDF) vectorization, unigram and bigram features, hand-crafted statistics and sentiment-based indicators, and several classification models such as Logistic Regression, Random Forest, XGBoost, and LightGBM. The issue of class imbalance is tackled using Synthetic Minority Over-sampling Technique (SMOTE) and compared. The original dataset of 7,489 samples was cleaned and narrowed down to 7,486 valid cases. Accuracy, Precision, Recall, F1 score, ROC-AUC and 5-fold cross-validation were used to evaluate the performance. There are experimental results to show that ensemble models are more effective compared to traditional linear classifiers. XGBoost performed best in the overall performance of 94.59% accuracy and F1-score of 0.8323. The hybrid-based feature fusion technique has a considerable improvement on the classification performance and does not sacrifice the level of interpretability and computational efficiency, which is why the framework is applicable to scalable mental health analytics services.
Authors - Armie E. Pakzad, Nathanael Adrian T. Cua, Louie T. Que, Alvin Josh T. Valenciano, Jana Johannes Valenzuela, Abbasali Pakzad Abstract - Emotional Support Conversation (ESC) seeks to lessen users’ emotional dis tress through sympathetic communication. Current approaches concentrate on comprehending present emotional states and combining support techniques to generate responses. But they fail to take into account an important factor: emotional trajectories (how users’ feelings change over time). Two people expe riencing the same feeling may need essentially different answers depending on whether they are in a therapeutic window (gradually improving), a depressed spiral (continuous hopelessness), or a crisis escalation (rapidly worsening). We propose TRAGEDY (TRAjectory-Guided Emotional Dialogue System), a sys tem that explicitly models clinical patterns and emotional trajectories in order to direct response creation. We present: (1) a trajectory encoder that records the temporal dynamics of emotion and intensity sequences; (2) a clinical pat tern detector that recognizes five psychologically grounded patterns (normal progression, therapeutic window, resistance pattern, depressed spiral, and crisis escalation); and (3) pattern-aware generation that bases responses on trajectories found. Experiments on the ESConv benchmark show that TRAGEDY provides interpretable trajectory insights while outperforming robust baselines, across standard generation metrics. Our approach opens new avenues for trajectory aware conversational AI and emphasizes the significance of temporal dynamics in emotional support.
Authors - Akhil P, Mallikharjuna Rao K. Abstract - Cloud storage platforms support diverse multimedia and col laborative workloads across organizations, yet conventional methods ne glect user behavior’s role in shaping access patterns. Privacy regulations prohibit centralized aggregation of interaction traces, while standard fed erated learning algorithms like FedAvg fail under statistical heterogene ity from varied user roles. This paper introduces FedPAE (Federated Per sonalized AutoEncoder), an unsupervised framework for behavior-aware user profiling in federated settings. FedPAE employs a shared global encoder for common patterns and private local decoders for individual adaptation, augmented by an Adaptive Fine-Tuning (AF) mechanism to mitigate encoder drift and preserve global semantics, without sharing any raw user data with the server. Evaluated on the CMU CERT benchmark and anonymized cloud storage logs, FedPAE surpasses FedAvg, FedProx, and FedPer in anomaly detection accuracy across all thresholds (e.g., F1 gains of 5–13% points over FedAvg across all precision thresholds), con f irming that the approach holds across heterogeneous client populations.
Authors - Yarragunta Babu , Challa Yuva Prasanthi, Vadapalli Sparjan, Sanagapati Venkata Siva Naga Sai Jayanth Abstract - Distributed systems rely on data replication across multiple nodes to ensure high availability, fault tolerance, and scalability. While replication improves system reliability, it also introduces temporary inconsistencies between primary and replica nodes during data propagation. This phenomenon, commonly referred to as consistency drift, occurs when distributed nodes maintain slightly different states before synchronization is completed. As distributed infrastructures grow in scale and complexity, consistency drift becomes increasingly significant due to network latency, workload variability, and communication overhead between nodes. Traditional synchronization mechanisms typically rely on static replication intervals or fixed update propagation strategies that do not adapt effectively to dynamic system conditions. Such approaches may allow drift to accumulate before synchronization occurs, resulting in delayed consistency and inefficient resource utilization. Managing consistency drift therefore becomes a critical challenge in distributed computing environments where maintaining accurate and synchronized data states is essential. This research addresses the problem of consistency drift in distributed systems by examining the factors that contribute to state divergence among nodes and exploring mechanisms for dynamic drift management. The proposed framework focuses on monitoring system behavior, including workload intensity, network latency, and node communication patterns, to regulate synchronization behavior more effectively. By enabling adaptive synchronization strategies that respond to real time system conditions, the framework aims to reduce drift accumulation and improve overall data consistency across distributed clusters. Effective management of consistency drift ultimately enhances system reliability, operational stability, and performance in modern distributed computing platforms operating under dynamic workloads.
Authors - Olutayo V. A., Agbele K. K., Ogundimu O. E., Dudu M. T. Abstract - As online shopping has become increasingly popular, companies must utilize social media to develop and improve customer experience. This study examined customer interaction sentiment regarding online shopping through automated systems to classify comments on social media sites like Twitter, Facebook, and Instagram. This research study compared three machine learning and natural language processing (NLP) techniques: Bidirectional Gated Recurrent Units (GRUs), Random Forests, and Naïve Bayes. Customer reviews were classified as positive, negative, and neutral, as well as analyzed for time-related patterns. The classification framework was constructed by using sentiment analysis, feature extraction, and data preprocessing techniques. Furthermore, model training and performance assessment were executed through Naïve Bayes and Support Vector Machines. Of all the models studied, the Bidirectional GRU had the best performance with an accuracy of 88.08 %. The results of this study help companies understand customer preferences better, and thereby refine their products, services, and marketing techniques.
Authors - Akbar Kushanoor, Sanjay K. Sahay Abstract - Traditional tree classification methods are inefficient, requiring tremendous effort, time, and labor. To address this, the primary objective of this research was to develop and implement a machine learning model that utilizes 3D Light Detection and Ranging (LiDAR) data, acquired via an unmanned aerial vehicle (UAV), for the accurate classification of tree species in the Philippines. Then, the collected data was pre-processed in preparation for the next portions of the methodology. Once completed, the features used in preparation for machine learning were extracted for the creation and training of the model. Ground truth data, validated by two licensed foresters, were used to ensure species accuracy, focusing on the five most abundant tree species in the dataset. Several machine learning algorithms were evaluated, with the XGBoost model achieving the best performance, reaching an overall accuracy of 85.63%, a mean class accuracy of 84.98%, and a Kappa accuracy of 81.57%. All producers’ accuracy exceeded 70%, indicating robust model reliability. Additionally, a user interface was developed to visualize the LiDAR data, tree attributes, and classification results. The findings demonstrate that LiDAR data obtained from UAVs can effectively be used for tree species classification in the Philippines, supporting forest inventory initiatives and reforestation efforts. Future work may include expanding the dataset, incorporating more species, and testing additional machine learning algorithms to further enhance classification accuracy.
Authors - Monir El Mounaoui, Kunale Kudagba, Mohamed Yassin Chkouri Abstract - This paper presents PricePulse, a web-based price comparison system that supports consumers with real-time multi-platform price analysis and AI-powered shopping insights. The system aggregates product data from Amazon, Flipkart, and Meesho via SerpAPI’s Google Shopping API and enriches results with recommendations generated by Google’s Gemini AI. Built on Next.js and Flask, PricePulse addresses gaps in the e-commerce ecosystem by eliminating manual price comparison across platforms. The system uses JWT-based authentication, maintains search history in SQLite, and provides an intuitive interface with React and Tailwind CSS. Evaluation shows average response times under 2 seconds and 95% accuracy in price extraction, demonstrating significant potential to help consumers make informed purchasing decisions and save on purchases.
Authors - Tiurida Lily Anita, Siti Nahdiah, Muslikhin Muslikhin, Mohd. Nor Shahizan Ali Abstract - Despite the importance of Allied Healthcare professionals in healthcare service delivery, low professional development opportunities, a high turnover rate, and a shortage of workers in India are some of the challenges that are affecting Allied Healthcare professionals’ retention. The purpose of this research is to explore the po tential of Internet of Things (IoT) solutions and Big Data analytics, coupled with infor mation and communication technology (ICT) as a solution to Allied Healthcare profes sionals’ retention strategies. The purpose of this paper is to propose a conceptual frame work that can be achieved by utilizing Internet of Things solutions coupled with Big Data analytics as a solution to Allied Healthcare professionals’ retention strategies by utilizing theories such as Technology Acceptance Model theory, Job Demands-Re sources theory, Social Exchange Theory, among others. The paper concludes that ICT is a resource that can be utilized to reduce job stress, enhance effective communication, and provide career opportunities for Allied Healthcare professionals; whereas Big Data analytics coupled with Internet of Things solutions can be utilized to predict potential risks that may affect Allied Healthcare professionals’ retention. The proposed concep tual framework offers a theoretical understanding of the digital revolution of human resource management practices in healthcare services.
Authors - Soji Binu Mathew, A. Hepzibah Christinal Abstract - Permanent Magnet Synchronous Motors (PMSMs) are commonly utilized in electric vehicle (EV) traction systems because of its high efficiency, power density, and reliability. Conventional field-oriented control (FOC) schemes require accurate rotor position and speed information, typically obtained from mechanical sensors, which increase cost and reduce system reliability. Sensor less control techniques based on observer theory have therefore gained significant attention. Among them, sliding mode observers (SMOs) offer strong robustness against parameter variations and external disturbances but suffer from chattering and noise sensitivity. This paper presents an advanced sensor less FOC strategy for PMSM drives using a super-twisting SMO (ST-SMO) for rotor position sensing and estimation of speed. The proposed approach employs a ST-SMO algorithm to achieve the convergence in finite-time while significantly reducing chattering effects. The observer is integrated into a standard FOC framework and evaluated under EV-relevant operating conditions, including low-speed operation and load transients. Comparative performance discussion demonstrates the suitability and the effectiveness of the proposed method for high-efficiency EV traction.
Authors - Mohammed Mudassir, Irene Joseph, Jyothi Mandala, Sandeep J Abstract - This study introduces a Bidirectional Long Short-term Memory based multichannel speech enhancement framework that operates in the short-time Fourier transform domain using time-varying complex spectral masking. The pro-posed approach predicts channel-specific complex masks, allowing adaptive frame-wise suppression of noise in reverberant and multi-noise environments. A comprehensive dataset was created using multiple noise sources, and experiments were carried out at different signal-to-noise ratios. The proposed method outperformed the Relative Transfer Matrix and Deep Multichannel Active Noise Control techniques in perceptual speech quality and intelligibility across all test conditions, indicating its potential for real-world speech enhancement applications.
Authors - Gauri P Nair, Vinaya V, Dona Sebastian, Kavitha K V Abstract - Reliable stock price forecasting remains challenging due to the noisy, nonlinear, and non-stationary characteristics of financial time-series data. Traditional statistical methods and deep learning models that rely solely on raw price data often struggle to capture short-term fluctuations and evolving market dynamics. To address these limitations, this study proposes a hybrid forecasting framework that integrates causal time-domain filtering, time–frequency feature extraction, and deep learning–based temporal modeling. The proposed approach employs Savitzky–Golay and Kalman filters to sup press high-frequency market noise while preserving important price trends in a causality-aware manner suitable for real-time forecasting. Localized spectral fea tures representing transient and time-varying market behavior are then extracted using the Short-Time Fourier Transform (STFT). These enhanced time-domain and frequency-domain features are combined and modeled using a Long Short Term Memory (LSTM) network, which effectively captures long-range depend encies and nonlinear temporal patterns in financial data. The framework is evaluated using standard performance metrics, including RMSE, MAPE, and R². Experimental results demonstrate that integrating causal filtering with STFT-based features significantly improves forecasting accuracy and robustness compared to baseline models, providing a reliable and practical solution for short-term and multi-step stock price prediction.
Authors - Zubair Zaland, Mumtaz Begum Mustafa, Miss Laiha Mat Kiah, Hua-Nong Ting, Zuraidah M Don, Saravanan Muthaiyah Abstract - As digital marketing expands in Oman, many organizations struggle to transform large volumes of customer data into actionable insights. This study presents an AI-driven marketing intelligence framework designed for non-technical users, combining automated customer segmentation, sentiment analysis, and personalized recommendations. The framework employs an autoencoder-based feature extraction approach to capture key behavioral patterns, followed by K-Means clustering to define meaningful customer segments (Berahmand et al., 2024). A fine-tuned BERT model analyzes multilingual feedback in Arabic and English to assess customer sentiment (Manias et al., 2023). The framework was evaluated using 12 months of campaign data from 450 customers across multiple Omani businesses. Analysis revealed four distinct customer groups and an overall positive sentiment of +0.55. Controlled A/B experiments demonstrated that AI-guided campaigns outperformed traditional methods, increasing conversion rates by 27%, improving retention by 15%, and generating a threefold return on marketing spend. These results indicate that accessible AI tools can deliver measurable marketing benefits in emerging markets and provide a scalable solution for Gulf-region businesses.
Authors - B.Usha Rani, M.Sudhakar, A.Srivani, Y.Surya Praveen Abstract - The purpose of Diabetic Retinopathy Prediction is to use computer technology to identify early stages of retinal damage caused by diabetes. Since diabetic retinopathy can lead to blindness or permanent vision impairment if not treated in a timely manner, accurate and rapid diagnosis is vital. Recent tech niques for diagnosing diabetic retinopathy require an ophthalmologist to perform a manual examination of the eye’s retina with the use of fundus photography. The diagnostic process can be costly, time-consuming, and vary significantly from one person to another. A large percentage of diabetes patients live in rural areas, where it is difficult or impossible for them to have periodic screening by a diabetic specialist or receive healthcare services. There is a need to develop a solution to these problems, and the Diabetic Retinopathy Prediction System uses deep learning based techniques to analyze retinal fundus images and produce pre dictions regarding diabetic retinopathy. Analysis of the retinal fundus images will include preprocessing, feature extraction using CNNs, and automated classifica tion into diabetic retinopathy by degree and severity. This approach increases the accuracy and consistency of diabetic retinopathy diagnosis while minimizing the need for human input. The proposed system will allow for early identification of diabetic retinopathy in resource poor environments, support large scale screening programs and aid in clinical decision making by ophthalmologists. Additionally, the system has potential integration into mobile health systems and tele-ophthal mology networks. Experimental results indicate the proposed system is capable of accurately detecting diabetic retinopathy with high levels of specificity and sensitivity.
Authors - Bai B Mathura, Narra Dhanalakshmi Abstract - This paper presents a novel Reversible Data Hiding (RDH) method for dual images. First, secret data is converted into a binary sequence of equal length and then divided into shorter segments to control the amount of data embedded into each pixel. The embedding process uses two copies of the original image to distribute the data, reducing the impact on each image while maintaining overall image quality. During recovery, the original image is restored by averaging the pixel values at corresponding locations in the two stego images, while the embedded data is recovered through a reverse process. Experimental results on grayscale images demonstrate that the method maintains good image quality, achieving a high Peak Signal-to-Noise Ratio (PSNR) across different embedding levels while ensuring accurate recovery of both the secret data and the original image.
Authors - Priyanka Khalate, Satish S. Banait, Chandrakant Kokane, Dnyanada Shinde, Madhumati Pol, Pravinkumar M. Sonsare Abstract - The emerging use of digital deepfake technology is creating a myriad of obstacles in verifying the authenticity of digital media. Most of today’s detection methods yield satisfactory results when applied to clean samples of content, however, they are still susceptible to adversarial perturbations specifically created to bypass these detection methods. The current research paper introduces DC-DAFDN, a dual-stream architecture for detecting fraudulent digital content, which fuses frequency-domain analysis using the Discrete Cosine Transform (DCT) with Space-Attention Mechanisms. The current architecture uses adversarial training to develop more robust features. The proposed model uses EfficientNet-B4 as a backbone, augmented with Spatial Reduction Attention Blocks and Forged Fea tures Attention Modules to detect manipulation artifacts in the spatial domain, while the parallel DCT stream analyzes inconsistencies in the frequency-domain. Through an adversarial training procedure using Fast Gradient Sign Method (FGSM)-induced adversarial perturbations, the model learns robust feature sets that are resistant to evasion attacks. When evaluated on Face-Forensics++ dataset, DC-DAFDN significantly improves upon the original Dual Attention for Deepfake Detection Network (DAFDN) in terms of adversarial robustness. When attacked with large adversarial perturbations (e.g., FGSM with ϵ ranging from 0.1 to 0.25), the DC-DAFDN architecture maintained greater than average accu racy enhancements from +2.74% up to +3.61%, for an average accuracy increase of +3.36%, for the tested att, from all strengths. Our findings suggest that fusing frequency-domain analysis with adversarial training provides measurable improvement in the model’s robustness to adversarial attacks and simultaneously preserves the detection capabilities of the dual-attention method.
Authors - Cristian Castillo-Olea, Clemente Zuniga Gil, Angelica Huerta Abstract - Question paper preparation in educational institutions is conventionally manual and time-consuming, often generating question papers of uneven difficulty and less diversity. This project solves the problem of automatic question paper generation from voluminous academic content available in multiple formats. The motivation for this work is reducing human effort and enhancing efficiency, ensuring fair and balanced assessment generation, while supporting modern digital learning environments. Input content, in the form of text documents, portable document files, presentation slides, images, audio recordings, and video lectures, forms the bedrock of the proposed system; first, it gets preprocessed into a unified textual format through document parsing, optical character recognition, and speech-to-text techniques. Natural language processing approaches like sentence segmentation, tokenization, stop word removal, and extraction of key concepts are subsequently applied on the meaningful and relevant identification of the contents. It follows a hybrid approach relying on the Transformer architecture: a classification model that assesses the importance of a sentence, relevance of concepts, and difficulty level; and a generation model providing question types such as multiple choice, short answer, long answer, case studies, reasoning, fill-in-blanks, and programming. The proposed model goes through training and fine-tuning using publicly available datasets of question-answer pairs and pre- processed information in textbooks. In the experimental results, the proof of efficiency by the proposed approach is shown in generating accurate and diverse question papers with high relevance. Such an approach would definitely ensure much better outcomes for the question papers and the assessment.
Authors - Y. C. A. Padmanabha Reddy, Panigrahi Srikanth, Kavita Goura Abstract - Advances in Artificial Intelligence, Machine Learning and Internet of Things technologies have enabled wearable devices to sense as well as process and respond to human behaviour in real time. While most wearable devices today are used for health and fitness tracking. Many people face communication challenges such as language barriers, difficulty understanding emotions or social cues, social anxiety and accessibility issues for individuals with hearing or speech impairments. Existing systems often collect data but fail to provide meaningful, real-time assistance during actual human interactions. This research paper presents a literature-based study on AI powered wearable devices designed to support and enhance human communication. The research papers are focusing on intelligent wearables that use multimodal sensors such as microphones, cameras and sensors. These systems apply AI techniques to interpret speech, gestures, facial expressions and emotional signals in real time. The wearable devices considered include everyday consumer-oriented systems such as smart eyewear that provides audio visual assistance and wrist worn wearables that offer haptic feedback. The key focus of this study is to examine how such devices can deliver subtle, real-time support through visual prompts, audio cues or vibrations to improve conversational awareness and user confidence. The expected outcome is to identify current capabilities, practical limitations and design considerations for developing human centric wearable technologies that move beyond passive tracking toward meaningful communication support.
Authors - Anisha Panja, Ranjita Kumari Dash, Biswajit Sahoo Abstract - Singer identification is a challenging task because of pitch and me lodic variations, tempo, vibrato, and adaptive singing styles. This paper propos es a novel approach towards singer identification and classification by adapting a model originally meant for speaker recognition. Specifically, this work utiliz es vector representations extracted from a pretrained Speech Brain Emphasized Channel Attention, Propagation and Aggregation in Time Delay Neural Net work (ECAPA-TDNN) model. The research pipeline processes a custom curated dataset of four prominent Indian playback singers into fixed, 8 second audio clips, with mono channel sampled at 16 kHz and exported as wav files. The Speech Brain Emphasized Channel Attention, Propagation and Aggrega tion (ECAPA) encoder transforms these labelled clips into fixed embeddings which are unique vector representations of voice characteristics of each audio clips. A suite of classical machine learning classifiers is trained on these em beddings. The study evaluates four of them namely, Logistic Regression, Sup port Vector Machines, Random Forests, and a Multi-Layer Perceptron (MLP). The MLP achieved the highest accuracy of 99.38% on held-out test data. Sup porting this result, both confusion matrix analysis and t-SNE projection clearly demonstrate clear cluster separation based on individual singer identities. These findings thus collectively validate that ECAPA embeddings contain sufficient identity-bearing structure on a singing voice. This analysis thus concludes that adaptation of speaker recognition models with appropriate classifiers is a great ly effective and efficient approach for singer identification.
