G Pulla Reddy College of Engineering & Technology

The most important research field in the current period is facial recognition and detection. Furthermore, facial expression recognition applications are essential in various research areas such as criminal investigation, security, data base management systems (DBMS), innovative card application, and video surveillance systems. A criminal investigation is an important area of study. Thus, in today’s world, crime activities are fast expanding in response to psychological trauma. As a result, applying deep learning (DL) algorithms for face expression identification and crime activity monitoring is a growing area of investigation. In terms of pre-processing, feature extraction, and recognizing diverse facial emotions, DL produces better results. As a result, the current study seeks to offer a revolutionary intelligent Strawberry-based convolution neural network (SbCNN). When applied to the Kaggle face expression database, the created SbCNN technique performed best. Face verification and criminal face detection can be accomplished using this approach and the optimization fitness function. Even though the established design improves the robustness of feature extraction and classification, it is a time-consuming process.

The rapid growth of smart cities, powered by Internet of Things (IoT) technologies, demands robust, energy-efficient, and scalable cybersecurity solutions. As urban-scale systems increasingly depend on massive networks of sensors and edge devices, ensuring secure and sustainable communication becomes a critical challenge. This research presents a framework for Sustainable Cybersecurity Solutions, emphasizing Decentralized and Resource-Efficient Architectures for energy-optimized security in smart environments. Research proposes an Intelligent and Secure Edge-Enabled System (ISEC) model that integrates Green IoT, edge computing, and artificial intelligence (AI) to achieve secure, low-latency data transmission. It uses the smart city IoT and edge network dataset, applying Z-score normalization during preprocessing to standardize features and improve model performance, followed by Linear Discriminant Analysis (LDA) for feature extraction to maximize discriminative information and reduce dimensionality, thereby improving detection accuracy. The framework employs deep learning-based Termite Colony Optimizer-Driven Stacked Bidirectional Long Short-Term Memory (TCO-Stacked BiLSTM) to identify optimal routes and predict potential threats, enabling real-time threat detection and mitigation across the network. Edge computing decentralizes processing closer to IoT nodes, minimizing latency and reducing energy use. The ISEC model leverages this structure to avoid centralized bottlenecks, therefore embodying a resource-efficient architecture that balances security, computation, and power consumption. Low-powered sensors are supported through optimized routing protocols and lightweight security mechanisms, which reduce processing load and communication overhead. Experimental analysis shows the proposed TCO-Stacked BiLSTM model achieves accuracy, precision, recall, and F1-score ranging between 94% and 98%, along with reductions in energy consumption, latency, improvements in throughput, and significant enhancements in reliability, demonstrating efficient, low-latency, and resource-conscious performance for smart city IoT networks. These results confirm the efficiency and scalability of the model. Overall, the proposed solution provides a sustainable, secure, and resource-conscious framework, well-suited to the demands of modern smart cities and future urban digital infrastructures.

Data center security is enhanced with the rapid deployment of AI and ML in cybersecurity, yet energy use and carbon emissions have also increased with this development. In large-scale data center operations, this dual issue emphasizes the need for techniques that advance carbon neutrality while ensuring strong security. This research aims to design a sustainable cybersecurity framework that enhances computational performance while reducing environmental impact through energy-efficient modeling and optimization. This hybrid approach proposed an Oneclass Support Vector Based Bidirectional Snow Geese Algorithm (OSV-Bi-SGA). Carbon aware-cybersecurity traffic datasets are preprocessed through data cleaning, Z-score normalization, and categorical encoding to ensure robust input for modeling. Feature extraction is conducted using principal component analysis (PCA). The proposed OSV-Bi-SGA method is integrated with Bidirectional Long Short-Term Memory (BiLSTM), which captures temporal bidirectional dependencies in traffic sequences. Oneclass support vector machine (OSV) identifies anomalies when only normal class data is available. Snow Geese Algorithm (SGA) enhances parameter optimization, reducing energy cost while maintaining performance. The suggested OSV-Bi-SGA model achieved a high precision (99.42 %), recall (99.24 %), and F1-score (99.32 %), while reducing energy consumption and carbon footprint compared to baseline models. The research demonstrates that integrating evolutionary optimization with deep learning (DL), machine learning (ML) and anomaly detection can balance high-performance cybersecurity with reduced environmental impact. The OSV-Bi-SGA framework provides a promising pathway for sustainable and carbon-neutral data center security operations.

With the rapid growth of Internet technology and social networks, the generation of text-based information on the web is increased. To ease the Natural Language Processing (NLP) tasks, analyzing the sentiments behind the provided input text is highly important. To effectively analyze the polarities of sentiments (positive, negative and neutral), categorizing the aspects in the text is an essential task. Several existing studies have attempted to accurately classify aspects based on sentiments in text inputs. However, the existing methods attained limited performance because of reduced aspect coverage, inefficiency in handling ambiguous language, inappropriate feature extraction, lack of contextual understanding and overfitting issues. Thus, the proposed study intends to develop an effective word embedding scheme with a novel hybrid deep learning technique for performing aspect-based sentimental analysis in a social media text. Initially, the collected raw input text data are pre-processed to reduce the undesirable data by initiating tokenization, stemming, lemmatization, duplicate removal, stop words removal, empty sets removal and empty rows removal. The required information from the pre-processed text is extracted using three varied word-level embedding methods: Scored-Lexicon based Word2Vec, Glove modelling and Extended Bidirectional Encoder Representation from Transformers (E-BERT). After extracting sufficient features, the aspects are analyzed, and the exact sentimental polarities are classified through a novel Positional-Attention-based Bidirectional Deep Stacked AutoEncoder (PA_BiDSAE) model. In this proposed classification, the BiLSTM network is hybridized with a deep stacked autoencoder (DSAE) model to categorize sentiment. The experimental analysis is done by using Python software, and the proposed model is simulated with three publicly available datasets: SemEval Challenge 2014 (Restaurant), SemEval Challenge 2014 (Laptop) and SemEval Challenge 2015 (Restaurant). The performance analysis proves that the proposed hybrid deep learning model obtains improved classification performance in accuracy, precision, recall, specificity, F1 score and kappa measure.

This study introduces an integrated experimental and finite element analysis (FEA) simulation methodology for improving the turning process of Inconel 825 using tungsten carbide (WC) cutting tools. The research presents an innovative framework that integrates infrared thermal imaging with numerical simulations to examine transient temperature profiles and cutting forces across different machining settings. This study systematically examines the effects of feed rate, cutting speed, and depth of cut on thermal and mechanical responses, employing an L9 orthogonal array for experimental design, in contrast to usual investigations. This research’s primary innovation is the exact monitoring of interface temperatures with an infrared thermal camera, yielding precise thermal data despite the difficulties posed by expensive materials and real-time heat dissipation assessment. The FEA simulations performed in Abaqus FEA utilize an elastoplastic material model exhibiting nonlinear behavior, effectively capturing yielding in both tension and compression. The results demonstrate a robust connection between experimental and numerical findings, with cutting force predictions differing by less than 5%. The research indicates that raising the cutting speed lowers cutting forces while influencing temperature patterns in a non-linear manner. The research underscores the significance of WC inserts in augmenting heat dissipation and promoting machining stability. The proven FEA framework provides a dependable prediction instrument for optimizing machining settings, hence enhancing process control and precision manufacture of high-strength alloys.