G. Pullaiah College of Engineering and Technology

In the Internet of Things (IoT) era, energy efficiency and reliability of static random-access memory (SRAM) play a critical role in achieving low-power and robust system performance. This paper introduces a novel SRAM topology that mitigates read and half-select disturbances while simultaneously enhancing read static noise margin (RSNM) and write static noise margin (WSNM). The proposed design improves RSNM through a singleended isolated read path, and WSNM is strengthened using a power-rail switching scheme, in which the inverter's VDD is temporarily connected to ground. To further reduce power consumption, transistor stacking is incorporated in both inverters and write paths, eliminating read bitline leakage and removing the need for dedicated write bitlines. Compared to prior SRAM architectures, the proposed design achieves up to 1.95 & times; improvement in RSNM and 1.47 & times; in WSNM, while reducing read power, write power, and leakage power by at least 17.13%, 25.81%, and 34.78%, respectively. Moreover, under severe process variations, the proposed design demonstrates significantly lower variability in RSNM, WSNM, and leakage power. Further, the superiority of the proposed SRAM design is established through its electrical quality metric, a key measure of overall performance. All evaluations were performed using Synopsys HSPICE with 16-nm CNTFET technology, considering a 64 & times; 64 SRAM array.

Effective and secure attendance management is absolutely needed in both educational institutions and workplaces, but traditional approaches of using manual registers, RFID cards and biometrics that contact the participant have issues of proxy attendance, hygiene and operational inefficiency. The face recognition operated vision-based contactless attendance systems would be a good alternative, but their accuracy reduces in the real-life experience of face mask usage, partial coverage, changes in light, and presentation attack. To overcome these constraints, this paper proposes a vision-based contactless system of attendance that combines dense geometry of face mesh with iris landmark and texture scan. The suggested system uses stable facial geometry representation based on real-time three-dimensional facial landmarks to capture stable facial geometry features and the iris segmentation and landmark detection module, which is based on deep learning, to extract discriminative ocular features. Multimodal feature fusion-based liveness-aware verification is applied to attendance decision making. It is experimentally proven that the given system has recognition accuracy 98.4, and false acceptance/ false rejection are 1.1 and 1.5, respectively, which is better than face-only and iris-only. Moreover, the system has the accuracy of 96.9% in mask conditions and 94.9% average presentation attack, so there is no doubt that the system is robust, secure, and can be used in real-time contactless attendance applications.

Complex team sports require sophisticated, rapid-dynamic decision-making schemas that adjust to tactical situations and adversarial behaviors. This makes tactically cohesive, field-applicable policies difficult. TACT-RLNet, a modular reinforcement learning framework based on context-based tactical concepts for active coordination and real-time adaptation in multi-agent team sports, addresses these issues. Five key modules comprise TACT-RLNet. (1) CMATA-credited attribution specificity by counterfactually calculating each player’s reward alignment and accountability contribution; (2) Phase-Aware Tactical Encoder (PATE) captures strategic game phases like attacking or defending and embeds tactical context to improve policy coherence under temporal shift (3) Opponent Tactical Adaptation Layer (OTAL)—real-time data-driven strategy model upgrades and progressive policy modification; 4. Communication-Constrained Curriculum Trainer (C3T)—creates real conditions for training policy under limited communication constraint for good coordination; and 5. Spatial Threat Mapping and Reward Shaping Engine (STMRSE)—pressure zone modeling and reward shaping based on spatial influence to support behavior organization like marking and pressing Performance enhanced 25

Knee osteoarthritis (OA) is a popular and progressive degenerate joint disease and require a predictor of risk at the primary stage to predict drug treatment and propose an early intervention. The current methods of deep learning are not sufficient enough to deal with integrating the heterogeneous modality component, modeling of irregular temporal follow ups, and capture of the fine grained variations of pathological types. In order to address such limitations, this paper suggests a new multimodal approach named, RAD-OAEnsemble which introduces three main innovations. First, a novel Cross Bi-modal Heterogeneous Graph Fusion Network (CBHGF-Net) is introduced for the multi-scale semantic fusion of Radiographs and Electronic Health Records (EHRs). Second, a Modified DenseNet with Multi-scale Attention Criss-Cross Blocks (MACC-Blocks) is presented to extract the anatomical features of knee OA patterns by utilizing dilated convolution, criss-cross attention, residual attention gating. Third, a Dual-Stream Tranfo-Convo (DTC) Deep Temporal Embedding module is proposed that consists of a Time-Warped Convolutional Blocks and a Temporal Decay Encoded Informer for the knee OA risk prediction. The suggested approach delivered better outcomes with accuracy of 99.275%, precision of 98.175%, specificity of 99.165%, sensitivity of 98.95%, and F1-score of 98.175% in having considerably a low rate of false positives (1.875%) and false negatives (2.175%). The findings support the relevance of the suggested framework in providing resilient, time-sensitive, and explainable OA progression predictions by multimodal combination and high-fidelity feature learning.

High population mobility and imbalances of housing markets are the critical challenges to sustainable urban and regional planning. The traditional forecasting models tend to focus on demographic change and housing pressure as independent aspects, hindering their capacity to capture interdependence of the two aspects. The current paper suggests a multi-task spatio-Temporal learning model that closely predicts the migration pressure and housing stress with the help of publicly available macroeconomic and demographic statistics on a national scale. Composite indices that capture the pressure of migration, the tightness of houses, and the affordability risk are obtained and as such, they are sourced and fed into a common Long Short-Term Memory (LSTM) architecture with two prediction heads. Experimental analysis using a country-level panel data has good predictive power, especially among housing stress predictions, where the model has a mean absolute error of 0.718, root mean squared error of 0.888, and a positive coefficient of determination (R2 = 0.269), which is better than statistical and single-task baselines. The prediction of migration pressure is still competitive with an MAE of 0.779 and an RMSE of 1.080. Further experiments involving cross country transfer reveal that the representations learnt in common can be used with predictive power in varying national states. These results demonstrate the usefulness of integrated modeling in the context of migration-housing interactions and give a facts-based basis on which to anticipate the appearance of new housing stress patterns at the national level at an early stage.