Pose variation in facial imagery presents a persistent challenge for automated face recognition systems, particularly in uncontrolled environments such as surveillance, access control, and mobile device authentication. This paper introduces an approach based on Conditional Generative Adversarial Network (cGAN) for synthesizing photorealistic frontal views from single profile images. The proposed architecture concatenates a spatially replicated noise vector with the input profile, enabling generation diversity while retaining subject identity. A composite loss function integrating adversarial, L1, and L2 losses is employed to enhance both global realism and pixel-level fidelity. The model is trained on a custom dataset comprising 4,682 images of 44 subjects, each with a single frontal view and multiple side profiles. Training is performed incrementally to improve stability and convergence. Qualitative results indicate that the method produces visually convincing frontal images with preserved identity details. This work establishes a foundation for future extensions involving perceptual loss, identity-preserving regularization, and large-scale evaluations.

This paper proposes a facial expression recognition (FER) based system for predicting mental health in humans. The system is designed with four key components. The first component involves image acquisition and preprocessing, where the facial region of interest is extracted from the captured image. The second component focuses on extracting features from the facial region and classifying facial expressions. This is achieved through multi-scale and multi-deep feature representation approaches. Specifically, a multi-scale convolutional neural network (CNN) architecture and a multi-deep bilinear CNN architecture are employed to extract discriminative and complementary features for accurate classification into facial expression categories. The third component enhances the performance of the FER system by utilizing scores generated by the multi-deep CNN models. The fourth component predicts and analyzes mental health scores based on the FER classification outputs. These scores are computed statistically using established studies on mental health from the physiological literature. Experimental results reveal that the FER model achieves higher detection rates for negative emotions and lower detection rates for positive emotions. The system’s performance is evaluated on two benchmark databases: the Karolinska Directed Emotional Faces (KDEF) and AffectNet, a database for facial expression, valence, and arousal analysis in natural settings. A comparative analysis with state-of-the-art methods reveals F1-scores of 82.72% and 65.03% for the KDEF and AffectNet databases, respectively. These results demonstrate the superiority of the proposed system in both facial expression classification and mental health prediction, showcasing its effectiveness in addressing this critical application domain.

Protein methylation is a crucial post-translational modification (PTM) responsible for many diseases and accurate prediction of the methylation site is important for understanding the molecular mechanism of the disease. The models have been successful in capturing contextual dependencies in protein sequences, with deep learning models, specifically those based on the Transformer architecture and Multi-Head Attention, exhibiting good performance. However, most existing techniques rely on the sequence-only or hand-crafted features and are unable to capture biochemical properties and positional patterns, thereby limiting cross-species generalization and prediction accuracy. To cater for such demands, PLM-ArgMe is presented that is based on a symmetry-sensitive Transformer framework using context-aware ESM-2 residue embeddings, which is mapped through a novel Bio-Symmetric Mirrored Sinusoidal Encoding (BSMSE) strategy to address the biological symmetry hypothesis of arginine methylation. ESM-2 encodes evolutionary and structural context, while biochemical representations are enhanced by physicochemical features. A symmetry-aware positional encoding strategy and bidirectional multi-head self-attention are used to model structural, sequence-level, and feature-level dependencies. The proposed framework, PLM-ArgMe, achieves prediction accuracies of 90.91%, 93%, 87.44%, and 87.22% on Chimpanzee, Rat, Human, and Mouse datasets, respectively. When trained and evaluated on a combined multi-species dataset, the model attains an overall accuracy of 88.41%. The results reveal good generalization on a variety of datasets and suggest that PLM-ArgMe is a robust method for arginine methylation site prediction.

Precision agriculture productivity and efficiency are increased by integrating cross-industry technologies through Artificial Intelligence techniques. This work possesses a data-driven design and analysis of soil properties for an Artificial intelligence-based Safron cultivation prediction system in an agroinformatics framework. Safron is an expensive, cash-crop-growing spice, which has a significant impact on the crop industry’s economics. Several agronomic properties are essential for safron cultivation, but the soil’s agronomic properties are particularly crucial. In this work, three soil agronomical properties, such as (i) chemicals, (ii) micronutrients, and (iii) potential of hydrogen and electrical conductivity properties, are considered. The implementation of this work has four components. The first component involves data-driven techniques for these three different sets of soil properties. The second component consists of the data analysis of the collected soil properties, where statistical-based methods, such as descriptive statistics, factor analysis, and structural equation modelling techniques, have been employed. The third component utilizes machine learning techniques and strategies to verify the outcomes of structural equation modelling, followed by building a prediction system using unsupervised neural network-based models to derive predictions for the Safron Crop Prediction System, as outlined in the fourth component. Here, the experiments are carried out using observational and inferential-based studies for artificial intelligence-based saffron cultivation in the agroinformatics framework.

This work presents a comprehensive sustainable sentiment analysis system utilizing textual data, designed within a structured client-server architecture for real-time deployment. The system integrates dual feature representations Bag-of-Words (BoW) and Term Frequency-Inverse Document Frequency (TF-IDF) whose prediction scores are combined through a parameter-free score-level fusion strategy. The implementation of the proposed system consists of five major components. The first component involves the acquisition of textual data from various sources, followed by rigorous text preprocessing to eliminate noise and enhance data quality. The second component focuses on feature extraction, ensuring that the extracted features not only reduce computational overhead but also retain high discriminative capability to effectively represent sentiments. The third component involves the development of sentiment classification models to categorize textual data based on sentiment polarity. The framework is evaluated across binary (2-class), ternary (3-class), and fine-grained (13-class) sentiment and emotion datasets. To address class imbalance in the 13-class setting, LLM-based data augmentation is incorporated during training. Following model development, the fourth component involves performance optimization, where the best-performing feature models are selected, and classifier parameters are fine-tuned to maximize accuracy and efficiency. Finally, in the fifth component, the optimized and scalable sentiment analysis models for both three-class and thirteen-class sentiment categories are deployed to a server environment for sustainable real-time sentiment classification tasks. Experimental results demonstrate that the proposed architecture maintains stable predictive performance while supporting controlled vocabulary sizes for efficient deployment. The system design emphasizes thin-client interaction, stateless REST-based inference, and centralized model management, ensuring practical applicability in real-world environments, making the system suitable for large-scale implementation in various domains such as customer feedback analysis, social media monitoring, and opinion mining.

This research examines the integration of educational systems with electronic learning, focusing on how information and communication technology (ICT) may impact children’s mental health. This research highlights the impact of ICT-integrated education technology on children’s mental well-being, encompassing enhanced access to mental health resources, facilitation of remote counselling, and promotion of social connections. The proposed methodology outlines a structured three-stage approach, including data collection, statistical techniques, machine learning techniques, and hypothesis testing. In the first stage of the method, two standard questionnaires, the Perceived Stress Questionnaire (PSQ) and the Depression, Anxiety, and Stress Scale (DASS-21), are used to assess children’s mental health. There are 468 children aged between ten and sixteen from an Indian school who have participated in answering the DASS21 and PSQ questionnaires. The data pre-processing techniques, followed by empirically statistical and machine learning-based methods, are employed in the second stage of the process. Then, several hypotheses are considered and validated based on the prediction of the mental health scores of children in the third stage of the study. Finally, the extensive experiments not only validate the objectives of this work but also build the best prediction model to calculate the scores and levels for depression, anxiety, and stress mental health problems for the children. This research highlights the importance of striking a balance between Education Technology and ICT usage, as well as offline activities, to foster children’s holistic development and mental well-being.

Saffron Crocus Sativus L. is among the most valuable cash crops globally, prized for its distinct color, aroma, and flavor, making it widely used in culinary and medicinal applications. However, the increasing global demand for saffron has also heightened the risk of adulteration, posing a significant challenge in its marketing and authenticity verification. Traditional methods for detecting saffron adulteration are expensive, time-consuming, and require specialized skills and advanced technology. This research proposes a machine vision-based, non-destructive saffron adulteration prediction system utilizing a Deep Learning (DL) based convolutional neural network model trained on RGB images of both genuine and adulterated saffron samples. The proposed model achieved an accuracy of 99.12 % , outperforming well-known transfer learning models such as ResNet50, InceptionNet, and VGG16. To enhance interpretability, explainable artificial intelligence techniques, including SHAP (SHapley Additive exPlanations) and LIME (Local Interpretable Model-agnostic Explanations), are employed to highlight the key features influencing the model’s predictions.

The search for exoplanets has advanced into the era of intelligent automation, yet most deep learning pipelines remain constrained to single-modality inputs or isolated views of astronomical data. We present PlanetNet-MMG, a novel multi-modal deep learning architecture that combines structured stellar metadata, raw lightcurve sequences, and graph-based relational context into a unified classification model. Our approach fuses three powerful encoders: a Tabular Transformer for domain-aware feature projection, a PatchGRU enhanced with a Vision Transformer (ViT) for learning fine-grained temporal patterns in segmented lightcurve patches, and a Graph PARE encoder that models inter-object similarity via a relational graph. Trained on a harmonized dataset derived from Kepler, TESS, and confirmed exoplanet archives, PlanetNet-MMG outperforms all state-of-the-art baselines, achieving a peak test accuracy of 90.4% and a class-averaged AUC of 0.973. Extensive experiments across 10-100 epochs and comparative evaluations against Astronet, ExoNet, OsbornNet, GCN (Lu), and ExoMiner confirm the effectiveness of our multimodal fusion. We further provide interpretability through attention overlays, t-SNE projections, and confidence histograms, reinforcing PlanetNet-MMG's transparency and reliability for scientific discovery in astrophysics.

This study introduces a multimodal sentiment analysis system to assess and recognize human pain sentiments within an Internet of Things (IoT)-enabled healthcare framework. This system integrates facial expressions and speech-audio recordings to evaluate human pain intensity levels. This integration aims to enhance the recognition system’s performance and enable a more accurate assessment of pain intensity. Such a multimodal approach supports improved decision making in real-time patient care, addressing limitations inherent in unimodal systems for measuring pain sentiment. So, the primary contribution of this work lies in developing a multimodal pain sentiment analysis system that integrates the outcomes of image-based and audio-based pain sentiment analysis models. The system implementation contains five key phases. The first phase focuses on detecting the facial region from a video sequence, a crucial step for extracting facial patterns indicative of pain. In the second phase, the system extracts discriminant and divergent features from the facial region using deep learning techniques, utilizing some convolutional neural network (CNN) architectures, which are further refined through transfer learning and fine-tuning of parameters, alongside fusion techniques aimed at optimizing the model’s performance. The third phase performs the speech-audio recording preprocessing; the extraction of significant features is then performed through conventional methods followed by using the deep learning model to generate divergent features to recognize audio-based pain sentiments in the fourth phase. The final phase combines the outcomes from both image-based and audio-based pain sentiment analysis systems, improving the overall performance of the multimodal system. This fusion enables the system to accurately predict pain levels, including ‘high pain’, ‘mild pain’, and ‘no pain’. The performance of the proposed system is tested with the three image-based databases such as a 2D Face Set Database with Pain Expression, the UNBC-McMaster database (based on shoulder pain), and the BioVid database (based on heat pain), along with the VIVAE database for the audio-based dataset. Extensive experiments were performed using these datasets. Finally, the proposed system achieved accuracies of 76.23%, 84.27%, and 38.04% for two, three, and five pain classes, respectively, on the 2D Face Set Database with Pain Expression, UNBC, and BioVid datasets. The VIVAE audio-based system recorded a peak performance of 97.56% and 98.32% accuracy for varying training–testing protocols. These performances were compared with some state-of-the-art methods that show the superiority of the proposed system. By combining the outputs of both deep learning frameworks on image and audio datasets, the proposed multimodal pain sentiment analysis system achieves accuracies of 99.31% for the two-class, 99.54% for the three-class, and 87.41% for the five-class pain problems.

This paper presents an advanced approach to a facial expression recognition (FER) system designed for robust performance across diverse imaging environments. The proposed method consists of four primary components: image preprocessing, feature representation and classification, cross-domain feature analysis, and domain adaptation. The process begins with facial region extraction from input images, including those captured in unconstrained imaging conditions, where variations in lighting, background, and image quality significantly impact recognition performance. The extracted facial region undergoes feature extraction using an ensemble of multimodal deep learning techniques, including end-to-end CNNs, BilinearCNN, TrilinearCNN, and pretrained CNN models, which capture both local and global facial features with high precision. The ensemble approach enriches feature representation by integrating information from multiple models, enhancing the system’s ability to generalize across different subjects and expressions. These deep features are then passed to a classifier trained to recognize facial expressions effectively in real-time scenarios. Since images captured in real-world conditions often contain noise and artifacts that can compromise accuracy, cross-domain analysis is performed to evaluate the discriminative power and robustness of the extracted deep features. FER systems typically experience performance degradation when applied to domains that differ from the original training environment. To mitigate this issue, domain adaptation techniques are incorporated, enabling the system to effectively adjust to new imaging conditions and improving recognition accuracy even in challenging real-time acquisition environments. The proposed FER system is validated using four well-established benchmark datasets: CK+, KDEF, IMFDB and AffectNet. Experimental results demonstrate that the proposed system achieves high performance within original domains and exhibits superior cross-domain recognition compared to existing state-of-the-art methods. These findings indicate that the system is highly reliable for applications requiring robust and adaptive FER capabilities across varying imaging conditions and domains.

Patient sentiment analysis is crucial for identifying issues, facilitating timely interventions, and enhancing healthcare quality. The management and analysis of pain are interconnected for delivering enhanced care. Self-assessed pain level evaluation is important in an intelligent healthcare framework for determining the most effective treatment. This study proposes an effective pain sentiment identification system based on analyzing patients’ facial expressions and EEG signals, considering diverse psycho-physiological traits in an efficient mobile healthcare system. The proposed system operates in four phases: (i) Patient facial regions and physiological dispositions are captured in the form of EEG signals. (ii) Extracted facial regions and EEG signals are analyzed through separate deep learning techniques to identify key features reflecting pain levels. (iii) Advanced feature tuning and representation techniques distinguish between low and high pain levels using deep learning models. (iv) Score level fusion enhances the performance of deep pain identification techniques within the architecture. The system’s performance was evaluated using the BioVid Dataset, and the results were compared to some existing well-known methodologies. Our proposed method has been rigorously evaluated through extensive experiments and has been shown to significantly outperform other state-of-the-art systems by 5

Saffron is an expensive cash crop obtained from the flowers of Saffron plant. Adulteration is one of the major menaces in the marketing of Saffron which needs to be addressed globally on priority basis. Multiple Saffron adulteration prediction systems have been proposed so far to predict the real and adulterated Saffron samples. However majority of these methods are based upon chemical and non-chemical approaches. The potential disadvantages associated with these approaches are that they are very expensive, complicated and need advanced technology and highly skilled experts. In this research work, a Saffron adulteration prediction system using machine learning has been proposed to predict the given Saffron samples into real or fake classes. The novelty of the system comes from leveraging the power of transfer learning networks to extract deep features from the images for improving the performance of machine learning models. The major advantages of the proposed system is that it is non-destructive, low-cost, and does not require an expert interference. Seven machine learning models have been used for classification including Gradient Boosting, XGB Classifier, AdaBoost, Random Forest, K-Nearest Neighbour (KNN), Support Vector Machine (SVM) and Decision Tree (DT) model. The SVM classifier outperformed all the other classifiers and gave an accuracy of 96.48 % . The performance of all the models was also compared with results of existing literature, and it was observed that the proposed feature extraction method and machine learning models performed well.

Patient sentiment analysis establishes an intricate relationship between pain management and sentiment analysis in delivering high-quality medical care. This work presents an efficient pain sentiment recognition system within a smart healthcare framework designed to assess patients' pain levels by analyzing their facial expressions. The proposed system is implemented in four distinct phases. First, facial regions are detected using efficient face-detection techniques. In the second phase, the extracted facial regions undergo feature computation using advancements in deep learning techniques, including end-to-end and pre-trained convolutional neural networks (CNN) to capture complex and discriminative facial features associated with pain emotions. In the third phase, a novel PainCapsule model is introduced, which evaluates pain intensity by analyzing both macro-and microfacial expressions. This phase also employs attention networks, feature tuning, and transfer learning techniques to optimize the system's performance. Finally, in the fourth phase, score fusion techniques are applied to the deep pain recognition models to enhance accuracy and robustness further. The system's effectiveness is rigorously evaluated using two benchmark video datasets: the BioVid Heat Pain Dataset and the Multimodal Intensity Pain (MIntPAIN) database. Extensive experiments and comparative analysis with existing state-of-the-art methods reveal that the proposed system achieves an F1-score of 65.51% for BioVid and 58.31% for MIntPAIN datasets, outperforming other pain recognition systems, demonstrating its potential to advance pain sentiment recognition within smart healthcare frameworks.

This study investigates the classification of both language and gender using speech signals from five distinct languages. A compre- hensive set of acoustic features is extracted, including Mel-Frequency Cepstral Coefficients (MFCC), Zero Crossing Rate (ZCR), Root Mean Square Energy (RMSE), statistical descriptors, and Chroma Short-Time Fourier Transform (Chroma STFT). Four ma- chine learning classifiers are evaluated on a merged dataset comprising SLR 41, 42, 43, 44, and SpeechOcean762. Experimental results reveal that features obtained from a 1D Convolutional Neural Network (CNN), when used with a Random Forest classifier, yield an F1-score of 98.49% for gender classification. For language identification, a Support Vector Machine (SVM) achieves an F1-score of 98.75%. In contrast, non-fine-tuned Wav2Vec embeddings produce lower performance, with F1-scores between 74% and 77%, while handcrafted acoustic features reach F1-scores of 96% for both tasks. These findings emphasize the continued effec- tiveness of traditional signal processing techniques in multilingual speech classification and highlight the necessity of fine-tuning when utilizing pre-trained deep learning models. The application of Principal Component Analysis (PCA) for feature reduction results in a slight decline in classification performance across all tested models.

This paper presents a face recognition system and identifies a person’s liveliness by recognizing his/ her facial spoof detection system. The implementation of the proposed system has three components: (i) real-time image data acquisition with human facial regions, (ii) extraction of the facial region from the input images, (iii) extraction of features by analyzing the facial pattern using both traditional and deep learning-based techniques for both face liveliness detection and recognition of a person, (iv) extensive experimentations have been performed to obtain the performances of both subject recognition and subject liveliness detection and comparing the performance with some existing state-of-the-art methods.

Blight disease poses a significant threat to agricultural output that results in large crop losses worldwide. Plant diseases must be promptly identified and managed to maintain crop health and maximise yields. This research presents a novel ensemble-based deep-learning model for plant blight disease detection, especially for agricultural applications. The suggested model uses convolutional neural networks (CNNs) for image recognition to accurately and automatically detect blight-affected areas in plant leaf images. An extensive dataset of plant leaf images was gathered to train and evaluate the model, including samples from both healthy and diseased plants. This ensemble-based deep learning model outperformed conventional deep learning and machine learning models in extracting characteristics that differentiated between plants affected by blight and those that weren’t. The proposed model (ResNet11) is a dependable and effective tool for on-the-spot disease detection in the field of potato, tomato and pepper, as demonstrated by experimental results that illustrate an accuracy of over 99 % for potato and pepper crops as a 3-class and 2-class problem respectively. Moreover, we get an accuracy of over 87 % for tomato plants as a 10-class problem.

A deep learning-based face anti-spoofing system has been proposed here. This work has been implemented in four segments. Firstly, an image preprocessing task is performed to extract the facial region. Then, the texture analysis of the facial region is performed to compute discriminant features. For this, a robust approach to deep learning techniques is needed, starting with defining some convolutional neural network (CNN) architectures for feature computation, followed by the classification of genuine vs. imposter face liveliness. The motivation of this work is to find both software- and hardware-based solutions to access biometric-based real-time systems through robust and vigorous face-liveness detection techniques. The recognition system's performances are further improved by image acquisition-challenging issues, image augmentation, fine-tuning, transfer learning, and the fusion of various trained CNN models. Finally, the above steps have been embedded in Raspberry Pi devices to build the system for real-time applications. The experimentation with two benchmark databases, NUAA and CASIA Replay-Attack, and comparing the performance with some well-known methods relating to the proposed system area show the proposed system's superiority.

In landslide-prone mountainous regions, accurate susceptibility prediction is crucial to mitigate dangers. Classical and probabilistic approaches have limited prediction capability. Hence, computational methods, specifically machine learning, are being employed to enhance accuracy. Our study aims to predict landslide susceptibility in the Kashmir Himalayas, specifically Muzaffarabad and the Azad Kashmir region. Initially, we established eleven distinctive features for prediction. We trained and tested seven machine learning models, comparing susceptibility predictions in two classes: susceptible and not susceptible. Evaluating classification performance, we achieved test accuracies ranging from 69.30 to 78.71

Protein methylation is a vital regulator of many biological processes at the post-translational level, and accurate prediction of protein methylation sites is essential for research and drug discovery. In this paper, we present a new method, namely RMSxAI, to predict the arginine methylation sites from primary sequences using machine learning algorithms and describe the predictions using explainable artificial intelligence (XAI) techniques. Leveraging experimentally validated methylated and unmethylated protein sequences from diverse organisms, we deduced several sequence features, encompassing physicochemical properties, amino acid composition, and evolutionary insights. Our results show that the proposed RMSxAI can predict protein methylation sites with high accuracy, bringing the F1 score up to 0.88 and overall accuracy up to 88.4

BACKGROUND: Patient sentiment analysis aids in identifying issue areas, timely remediation, and improved patient care by the healthcare professional. The relationship between pain management and patient sentiment analysis is crucial to providing patients with high-quality medical care. Therefore, a self-reported pain level assessment is required for the smart healthcare framework to determine the best course of treatment. OBJECTIVE: An efficient method for a pain sentiment recognition system has been proposed based on the analysis of human facial emotion patterns of patients in the smart healthcare framework. METHODS: The proposed system has been implemented in four phases: (i) in the first phase, the facial regions of the observation patient have been detected using the computer vision-based face detection technique; (ii) in the second phase, the extracted facial regions are analyzed using deep learning based feature representation techniques to extract discriminant and crucial facial features to analyze the level of pain emotion of patient; (iii) the level of pain emotions belongs from macro to micro facial expressions, so, some advanced feature tunning and representation techniques are built along with deep learning based features such as to distinguish low to high pain emotions among the patients in the third phase of the implementation, (iv) finally, the performance of the proposed system is enhanced using the score fusion techniques applied on the obtained deep pain recognition models for the smart healthcare framework. RESULTS: The performance of the proposed system has been tested using two standard facial pain benchmark databases, the UNBC-McMaster shoulder pain expression archive dataset and the BioVid Heat Pain Dataset, and the results are compared with some existing state-of-the-art methods employed in this research area. CONCLUSIONS: From extensive experiments and comparative studies, it has been concluded that the proposed pain sentiment recognition system performs remarkably well compared to the other pain recognition systems for the smart healthcare framework.