Search Results (14)

Emotion recognition in speech is essential for enhancing human–computer interaction (HCI) systems. Despite progress in Bangla speech emotion recognition, challenges remain, including low accuracy, speaker dependency, and poor generalization across emotional expressions. Previous approaches often rely on traditional machine learning or basic deep learning models, struggling with robustness and accuracy in noisy or varied data. In this study, we propose a novel multi-stream deep learning feature fusion approach for Bangla speech emotion recognition, addressing the limitations of existing methods. Our approach begins with various data augmentation techniques applied to the training dataset, enhancing the model’s robustness and generalization. We then extract a comprehensive set of handcrafted features, including Zero-Crossing Rate (ZCR), chromagram, spectral centroid, spectral roll-off, spectral contrast, spectral flatness, Mel-Frequency Cepstral Coefficients (MFCCs), Root Mean Square (RMS) energy, and Mel-spectrogram. Although these features are used as 1D numerical vectors, some of them are computed from time–frequency representations (e.g., chromagram, Mel-spectrogram) that can themselves be depicted as images, which is conceptually close to imaging-based analysis. These features capture key characteristics of the speech signal, providing valuable insights into the emotional content. Sequentially, we utilize a multi-stream deep learning architecture to automatically learn complex, hierarchical representations of the speech signal. This architecture consists of three distinct streams: the first stream uses 1D convolutional neural networks (1D CNNs), the second integrates 1D CNN with Long Short-Term Memory (LSTM), and the third combines 1D CNNs with bidirectional LSTM (Bi-LSTM). These models capture intricate emotional nuances that handcrafted features alone may not fully represent. For each of these models, we generate predicted scores and then employ ensemble learning with a soft voting technique to produce the final prediction. This fusion of handcrafted features, deep learning-derived features, and ensemble voting enhances the accuracy and robustness of emotion identification across multiple datasets. Our method demonstrates the effectiveness of combining various learning models to improve emotion recognition in Bangla speech, providing a more comprehensive solution compared with existing methods. We utilize three primary datasets—SUBESCO, BanglaSER, and a merged version of both—as well as two external datasets, RAVDESS and EMODB, to assess the performance of our models. Our method achieves impressive results with accuracies of 92.90%, 85.20%, 90.63%, 67.71%, and 69.25% for the SUBESCO, BanglaSER, merged SUBESCO and BanglaSER, RAVDESS, and EMODB datasets, respectively. These results demonstrate the effectiveness of combining handcrafted features with deep learning-based features through ensemble learning for robust emotion recognition in Bangla speech. Full article

Ambient Assisted Living (AAL) leverages technology to support the elderly and individuals with disabilities. A key challenge in these systems is efficient Human Activity Recognition (HAR). However, no study has systematically compared single-view (SV) and multi-view (MV) Human Activity Recognition approaches. This review addresses this gap by analyzing the evolution from single-view to multi-view recognition systems, covering benchmark datasets, feature extraction methods, and classification techniques. We examine how activity recognition systems have transitioned to multi-view architectures using advanced deep learning models optimized for Ambient Assisted Living, thereby improving accuracy and robustness. Furthermore, we explore a wide range of machine learning and deep learning models—including Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), Long Short-Term Memory (LSTM) networks, Temporal Convolutional Networks (TCNs), and Graph Convolutional Networks (GCNs)—along with lightweight transfer learning methods suitable for environments with limited computational resources. Key challenges such as data remediation, privacy, and generalization are discussed, alongside potential solutions such as sensor fusion and advanced learning strategies. This study offers comprehensive insights into recent advancements and future directions, guiding the development of intelligent, efficient, and privacy-compliant Human Activity Recognition systems for Ambient Assisted Living applications. Full article

In the wave of digitalization, organizations are increasingly focused on whether to prioritize digital culture or organizational strategy for the use of artificial intelligence (AI); there are mixed opinions, particularly when AI-augmented HRM draws attention as a tool for achieving sustainable organizational performance (SOP) in developing countries. This study aims to explore the influence of digital culture and organizational strategy on AI-augmented HRM and SOP, focusing on the mediating role of AI-augmented HRM in these relationships. To investigate the hypothesized relationships, 219 sample data were gathered from employees associated with HRM-oriented activities in Bangladesh, and SPSS 23 and AMOS software were used to test the SEM model. The results proved that digital culture has an insignificant effect and organizational strategy has a significant effect on AI-augmented HRM, and AI-augmented HRM has a substantial effect on SOP and partially mediates the relationship between organizational strategy and SOP. Based on the results, we infer that the successful implementation of AI-augmented HRM can lead to organizational sustainability in developing countries, where organizational strategy plays a pivotal role rather than digital culture. This research incorporates the resource-based view (RBV) and dynamic capabilities theories, which are crucial for the groundbreaking development of the research model. The results suggest that managers and responsible authorities should prioritize organizational strategy over digital culture when implementing AI-augmented HRM systems to ensure sustainability in developing countries. However, in the long run, organizations also need to concentrate on generating digitally favorable environments. Full article

Skin cancer poses a significant healthcare challenge, requiring precise and prompt diagnosis for effective treatment. While recent advances in deep learning have dramatically improved medical image analysis, including skin cancer classification, ensemble methods offer a pathway for further enhancing diagnostic accuracy. This study introduces a cutting-edge approach employing the Max Voting Ensemble Technique for robust skin cancer classification on ISIC 2018: Task 1-2 dataset. We incorporate a range of cutting-edge, pre-trained deep neural networks, including MobileNetV2, AlexNet, VGG16, ResNet50, DenseNet201, DenseNet121, InceptionV3, ResNet50V2, InceptionResNetV2, and Xception. These models have been extensively trained on skin cancer datasets, achieving individual accuracies ranging from 77.20% to 91.90%. Our method leverages the synergistic capabilities of these models by combining their complementary features to elevate classification performance further. In our approach, input images undergo preprocessing for model compatibility. The ensemble integrates the pre-trained models with their architectures and weights preserved. For each skin lesion image under examination, every model produces a prediction. These are subsequently aggregated using the max voting ensemble technique to yield the final classification, with the majority-voted class serving as the conclusive prediction. Through comprehensive testing on a diverse dataset, our ensemble outperformed individual models, attaining an accuracy of 93.18% and an AUC score of 0.9320, thus demonstrating superior diagnostic reliability and accuracy. We evaluated the effectiveness of our proposed method on the HAM10000 dataset to ensure its generalizability. Our ensemble method delivers a robust, reliable, and effective tool for the classification of skin cancer. By utilizing the power of advanced deep neural networks, we aim to assist healthcare professionals in achieving timely and accurate diagnoses, ultimately reducing mortality rates and enhancing patient outcomes. Full article

Higher levels of arsenic (As) and iron (Fe) in groundwater have been reported globally. This study aims to enhance our understanding of the role of naturally occurring dissolved Fe(II) in removing As from groundwater. Field experiments were conducted using five clay filters to investigate As and Fe removal from contaminated groundwater. The field results revealed a wide range of arsenic removal (7.3% to 80%) using the clay filters. The filter with the highest Fe concentration (14.5 mg/L) exhibited the highest As removal, while the lowest Fe concentration (2.2 mg/L) resulted in the lowest percentage of As removal. A direct correlation was observed between effluent As levels and the Fe/As molar ratio. An Fe/As molar ratio of 40 or more was identified as necessary to achieve effluent As concentrations below 50 µg/L. Laboratory batch experiments revealed that Fe(II) was more effective than Fe(III) in removing both As(III) and As(V) from contaminated groundwater. As(V) removal was consistently higher than As(III) removal, regardless of whether Fe(II) or Fe(III) was used. The results suggested that the oxidation of As(III) and the subsequent in situ formation of Fe(III) hydroxide were more efficient in As adsorption than direct Fe(III) treatment. The X-ray absorption fine structure (XAFS) analysis of the floc samples confirmed the dominant peaks of As(V), indicating that most of the As(III) oxidized to As(V) in the As(III)-Fe(II) system. The use of natural Fe(II) in groundwater, possibly supplemented with additional sources of Fe(II), is suggested as a promising, cost-effective, and efficient method for As(III) and As(V) removal. Full article

The ample amount of information from hyperspectral image (HSI) bands allows the non-destructive detection and recognition of earth objects. However, dimensionality reduction (DR) of hyperspectral images (HSI) is required before classification as the classifier may suffer from the curse of dimensionality. Therefore, dimensionality reduction plays a significant role in HSI data analysis (e.g., effective processing and seamless interpretation). In this article, a sophisticated technique established as t-Distributed Stochastic Neighbor Embedding (tSNE) following the dimension reduction along with a blended CNN was implemented to improve the visualization and characterization of HSI. In the procedure, first, we employed principal component analysis (PCA) to reduce the HSI dimensions and remove non-linear consistency features between the wavelengths to project them to a smaller scale. Then we proposed tSNE to preserve the local and global pixel relationships and check the HSI information visually and experimentally. Lastly, it yielded two-dimensional data, improving the visualization and classification accuracy compared to other standard dimensionality-reduction algorithms. Finally, we employed deep-learning-based CNN to classify the reduced and improved HSI intra- and inter-band relationship-feature vector. The evaluation performance of $95.21\%$ accuracy and $6.2\%$ test loss proved the superiority of the proposed model compared to other state-of-the-art DR reduction algorithms. Full article

Type 1 diabetes mellitus (T1DM) patients are a significant threat to chronic kidney disease (CKD) development during their life. However, there is always a high chance of delay in CKD detection because CKD can be asymptomatic, and T1DM patients bypass traditional CKD tests during their routine checkups. This study aims to develop and validate a prediction model and nomogram of CKD in T1DM patients using readily available routine checkup data for early CKD detection. This research utilized 1375 T1DM patients’ sixteen years of longitudinal data from multi-center Epidemiology of Diabetes Interventions and Complications (EDIC) clinical trials conducted at 28 sites in the USA and Canada and considered 17 routinely available features. Three feature ranking algorithms, extreme gradient boosting (XGB), random forest (RF), and extremely randomized trees classifier (ERT), were applied to create three feature ranking lists, and logistic regression analyses were performed to develop CKD prediction models using these ranked feature lists to identify the best performing top-ranked features combination. Finally, the most significant features were selected to develop a multivariate logistic regression-based CKD prediction model for T1DM patients. This model was evaluated using sensitivity, specificity, accuracy, precision, and F1 score on train and test data. A nomogram of the final model was further generated for easy application in clinical practices. Hypertension, duration of diabetes, drinking habit, triglycerides, ACE inhibitors, low-density lipoprotein (LDL) cholesterol, age, and smoking habit were the top-8 features ranked by the XGB model and identified as the most important features for predicting CKD in T1DM patients. These eight features were selected to develop the final prediction model using multivariate logistic regression, which showed 90.04% and 88.59% accuracy in internal and test data validation. The proposed model showed excellent performance and can be used for CKD identification in T1DM patients during routine checkups. Full article

Drought has exacerbated morbidity and mortality worldwide. Here, a time series study was conducted in northern Bangladesh to evaluate the impact of drought on selected causes of mortality during 2007–2017. Rainfall and temperature data from six meteorological stations were used to analyze drought and non-drought periods and to categorize mild, moderate, severe, and extreme drought based on the 3-month and 12-month Standardized Precipitation Index (SPI) and Standardized Precipitation Evaporation Index (SPEI). A generalized linear model with Poisson regression with log link, a negative binomial with log link, and a zero-inflated Poisson model were used to determine associations between drought severity and mortality. The SPI and SPEI produced slightly different analysis results. Compared with the SPEI, the SPI showed a stronger and more sensitive correlation with mortality. The relative risk for respiratory disease mortality was high, and Saidpur was the most vulnerable area. Health care expenditure was negatively associated with mortality. High temperatures during the drought period were associated with suicide-related mortality in Rajshahi. The impact of drought on mortality differed with small changes in climate. The findings of this study improve our understanding of the differences between the two most used drought indicators and the impact of drought on mortality. Full article

Since 2008, Bangladesh has had a school-based deworming programme to combat soil-transmitted helminth (STH) infection among school-aged children (SACs). Existing programmes have trouble reaching SACs, especially those out-of-school (OSCs). This study evaluated deworming coverage among school going children (SGCs) and OSCs in two Nilphamari sub-districts. It also evaluated community knowledge on STH control and deworming coverage in both areas for all SACs. Saidpur (intervention) and Kishoregonj (control) sub-districts, in Nilphamari, were surveyed in December 2019. The survey included SACs and their parents. Among SGCs, the intervention group (89.0%) had higher deworming coverage than the control group (75.5%). In the intervention group, 59.9% of OSCs received the deworming tablet versus 24.6% in the control group. Community involvement activities including door-to-door visits, courtyard gatherings, and miking benefited both SACs and their primary caregivers. SACs living in the intervention region, awareness of the last pill distribution date, and caregivers observing BRAC workers in action, were linked to SAC deworming coverage. Re-strategizing the deworming programme to include the OSCs is vital and suggests timely action. Building community awareness and periodic epidemiological assessment can further facilitate an improved drug intake. Full article

Chronic kidney disease (CKD) is one of the severe side effects of type 1 diabetes mellitus (T1DM). However, the detection and diagnosis of CKD are often delayed because of its asymptomatic nature. In addition, patients often tend to bypass the traditional urine protein (urinary albumin)-based CKD detection test. Even though disease detection using machine learning (ML) is a well-established field of study, it is rarely used to diagnose CKD in T1DM patients. This research aimed to employ and evaluate several ML algorithms to develop models to quickly predict CKD in patients with T1DM using easily available routine checkup data. This study analyzed 16 years of data of 1375 T1DM patients, obtained from the Epidemiology of Diabetes Interventions and Complications (EDIC) clinical trials directed by the National Institute of Diabetes, Digestive, and Kidney Diseases, USA. Three data imputation techniques (RF, KNN, and MICE) and the SMOTETomek resampling technique were used to preprocess the primary dataset. Ten ML algorithms including logistic regression (LR), k-nearest neighbor (KNN), Gaussian naïve Bayes (GNB), support vector machine (SVM), stochastic gradient descent (SGD), decision tree (DT), gradient boosting (GB), random forest (RF), extreme gradient boosting (XGB), and light gradient-boosted machine (LightGBM) were applied to developed prediction models. Each model included 19 demographic, medical history, behavioral, and biochemical features, and every feature’s effect was ranked using three feature ranking techniques (XGB, RF, and Extra Tree). Lastly, each model’s ROC, sensitivity (recall), specificity, accuracy, precision, and F-1 score were estimated to find the best-performing model. The RF classifier model exhibited the best performance with 0.96 (±0.01) accuracy, 0.98 (±0.01) sensitivity, and 0.93 (±0.02) specificity. LightGBM performed second best and was quite close to RF with 0.95 (±0.06) accuracy. In addition to these two models, KNN, SVM, DT, GB, and XGB models also achieved more than 90% accuracy. Full article

Climate change and severe emission regulations in many countries demand fuel and engine researchers to explore sustainable fuels for internal combustion engines. Natural gas could be a source of sustainable fuels, which can be produced from renewable sources. This article presents a complete overview of the liquefied natural gas (LNG) as a potential fuel for diesel engines. An interesting finding from this review is that engine modification and proper utilization of LNG significantly improve system efficiency and reduce greenhouse gas (GHG) emissions, which is extremely helpful to sustainable development. Moreover, some major recent researches are also analyzed to find out drawbacks, advancement and future research potential of the technology. One of the major challenges of LNG is its higher flammability that causes different fatal hazards and when using in dual-fuel engine causes knock. Though researchers have been successful to find out some ways to overcome some challenges, further research is necessary to reduce the hazards and make the fuel more effective and environment-friendly when using as a fuel for a diesel engine. Full article

Motor imagery (MI) from human brain signals can diagnose or aid specific physical activities for rehabilitation, recreation, device control, and technology assistance. It is a dynamic state in learning and practicing movement tracking when a person mentally imitates physical activity. Recently, it has been determined that a brain–computer interface (BCI) can support this kind of neurological rehabilitation or mental practice of action. In this context, MI data have been captured via non-invasive electroencephalogram (EEGs), and EEG-based BCIs are expected to become clinically and recreationally ground-breaking technology. However, determining a set of efficient and relevant features for the classification step was a challenge. In this paper, we specifically focus on feature extraction, feature selection, and classification strategies based on MI-EEG data. In an MI-based BCI domain, covariance metrics can play important roles in extracting discriminatory features from EEG datasets. To explore efficient and discriminatory features for the enhancement of MI classification, we introduced a median absolute deviation (MAD) strategy that calculates the average sample covariance matrices (SCMs) to select optimal accurate reference metrics in a tangent space mapping (TSM)-based MI-EEG. Furthermore, all data from SCM were projected using TSM according to the reference matrix that represents the featured vector. To increase performance, we reduced the dimensions and selected an optimum number of features using principal component analysis (PCA) along with an analysis of variance (ANOVA) that could classify MI tasks. Then, the selected features were used to develop linear discriminant analysis (LDA) training for classification. The benchmark datasets were considered for the evaluation and the results show that it provides better accuracy than more sophisticated methods. Full article

Binary (Chitosan-Cu(II), CCu) and Ternary (Chitosan-Alginate-Cu(II), CACu) composite materials were synthesized at variable composition: CCu (1:1), CACu1 (1:1:1), CACu2 (1:2:1) and CACu3 (2:1:1). Characterization was carried out via spectroscopic (FTIR, solids C-13 NMR, XPS and Raman), thermal (differential scanning calorimetry (DSC) and TGA), XRD, point of zero charge and solvent swelling techniques. The materials’ characterization confirmed the successful preparation of the polymer-based composites, along with their variable physico-chemical and adsorption properties. Sulfate anion (sodium sulfate) adsorption from aqueous solution was demonstrated using C and CACu1 at pH 6.8 and 295 K, where the monolayer adsorption capacity (Q_m) values were 288.1 and 371.4 mg/g, respectively, where the Sips isotherm model provided the “best-fit” for the adsorption data. Single-point sorption study on three types of groundwater samples (wells 1, 2 and 3) with variable sulfate concentration and matrix composition in the presence of composite materials reveal that CACu3 exhibited greater uptake of sulfate (Q_e = 81.5 mg/g; 11.5% removal) from Well-1 and CACu2 showed the lowest sulfate uptake (Q_e of 15.7 mg/g; 0.865% removal) from Well-3. Generally, for all groundwater samples, the binary composite material (CCu) exhibited attenuated sorption and removal efficiency relative to the ternary composite materials (CACu). Full article

This essay offers in-depth analysis of Guru Nanak’s works, collectively known as the Bābar-vāņī (“arrow-like utterances concerning Babur”), in the context of the memoirs of the first Mughal emperor Babur (1483–1530). It extends the number of works in the collection from a ‘fixed’ assemblage of ‘four’ to ‘nine,’ making it an open collection that dynamically responds to the specific questions raised by historians about Guru Nanak’s encounter with Babur. The resulting framework provides us with a fresh analytical gaze into the critical events related to Babur’s invasions of India and helps the novel readings of Guru Nanak’s verses shine through. It also examines how Guru Nanak’s voice of resistance was interpreted in the narratives produced by later generations. Departing from traditional views, the essay ends with a new understanding of the impact of the Bābar-vāṇī on the evolving Sikh conceptions of the relationship between spiritual and political powers. Full article