Bangladesh University of Business and Technology

We present ASL-HG, a comprehensive American Sign Language (ASL) image dataset designed to advance gesture recognition and assistive technologies. The collection contains 36,000 static images across 36 classes, covering the full English alphabet (A-Z) and digits (0-9). Data were captured from 10 volunteers in Mirpur, Dhaka, Bangladesh, with each participant contributing 100 samples per class, ensuring a balanced distribution across subjects, genders, and skin tones. Unlike many existing ASL datasets, ASL-HG explicitly distinguishes between the letter "O" and the digit "0″ by including the standard two-handed ASL "zero" sign used in practical alphanumeric communication. The dataset is released in two complementary forms: raw images with natural indoor and outdoor backgrounds, and a MediaPipe-processed version with hand-segmented crops and predefined 80-20 train-test splits. This design supports both custom pre-processing and immediate model training. ASL-HG is intended to serve as a benchmark resource for developing robust and fair ASL recognition systems, reducing communication barriers for deaf and speech-impaired users, and enabling broader research in gesture-based human-computer interaction.

Objective: Maintaining optimal vitamin D3 levels may play a key role in managing chronic low back pain (CLBP) and enhancing functional outcomes. This study aimed to compare the effects of health education and self-awareness booklets versus vitamin D3 supplementation combined with therapeutic exercises on pain intensity, vitamin D levels, and functional disability in individuals with CLBP.Material and Methods: This randomized clinical trial was conducted at two physiotherapy rehabilitation centers in Dhaka, Bangladesh, involving 84 participants with chronic low back pain (CLBP), randomly allocated into two equal groups. Group A received specific therapeutic exercises and a health education/self-awareness booklet, while Group B received vitamin D3 supplementation alongside the same exercises over a three-month period. Primary outcomes included pain intensity (assessed via the Brief Pain Inventory) and serum vitamin D3 levels. Functional disability, measured using the Roland-Morris Disability Questionnaire (RMDQ), served as the secondary outcome. Assessments were conducted at baseline and post-intervention, with a per-protocol analysis applied. Results: Both groups demonstrated significant improvements in pain intensity and functional disability following the intervention (p-value<0.001). However, Group B showed significantly greater reductions in disability scores (p-value<0.001), which corresponded with a marked increase in serum vitamin D3 levels (p-value<0.001).Conclusion: Both interventions were effective in reducing pain and disability in individuals with CLBP. However, the addition of vitamin D3 supplementation led to significantly greater improvements, highlighting its potential role in enhancing the efficacy of conservative CLBP management.

Purpose This study examines the implications of top executives' narcissism on auditor appointment and remuneration, focusing on the association between narcissistic leadership traits and audit outcomes in the context of an emerging economy, Bangladesh. Design/methodology/approach Using a hand-collected sample of 436 firm-year observations from publicly listed firms in Bangladesh over the period 2018–2021, this study employs panel regression to examine the impact of executive narcissism on auditor selection and audit fees. Endogeneity is addressed using a combination of entropy balancing, a two-stage least squares instrumental variable approach, and the Heckman two-stage selection model. Findings The findings indicate that narcissistic chief executive officers and chairs tend to appoint geographically proximate auditors while paying higher audit fees, potentially to exert greater influence over the audit process. This behavior is more pronounced in family firms and linked to faster audit report issuance, with no evidence of improved audit quality. However, Big 4 auditors weaken the observed effect. The results remain robust across alternative measures of executive narcissism and auditor–client distance, offering novel insights into how executive traits shape corporate governance and audit outcomes. Practical implications The study underscores the need for regulatory interventions to safeguard auditor independence and uphold audit quality, particularly in settings where narcissistic executives may prioritize personal influence over organizational integrity. It further highlights the importance of auditors to carefully assess risks associated with executive personality traits when determining audit fees and providing audit services. Originality/value This research is among the first to examine the interplay between executive narcissism and auditor–client geographic proximity within an emerging economy. By extending the literature on leadership traits, corporate governance and audit outcomes, the study provides novel insights of relevance to academics, policymakers and practitioners.

The abnormalities of the vocal folds cause voice disorders that affect speech and communication. The traditional diagnostic tools such as laryngoscopy and high-speed video endoscopy are subjective and time consuming and may be affected by inter-observer variation. In order to overcome these difficulties, we suggest a deep-learning-based paradigm of automatic vocal-fold segmentation and diagnosis of disorder on the Emory Vocal Folds dataset (approximately 2.9k of annotated images). We use preprocessing (cleaning, normalization and augmentation, class balancing), semantic segmentation as U-Net, and feature extractions as state-of-the-art CNNs (ResNet, DenseNet, VGG16, MobileNet and EfficientNet). The extracted features were used with both the end-to-end CNN classification and traditional machine learning classifiers (SVM, Random Forest, Gradient Boosting). It was experimentally demonstrated that EfficientNet was the most accurate at validation (98%), as compared to ResNet, DenseNet, and VGG16. Classical models also showed good results with an accuracy of 97% with random forest and 96 with SVM. The interpretability of the models was guaranteed through Grad-CAM, Grad-CAM++, Integrated Gradients, and Occlusion Sensitivity, and it was confirmed that the predictions paid attention to the parts of the vocal folds that are clinically relevant. These results indicate that deep learning can offer a robust, interpretable, and non-invasive algorithm for the early detection and diagnosis of vocal fold disorders thereby providing a possibility of real-time clinical decision support and mobile health diagnosis.

Automated vulnerability detection in source code is essential for modern software security, yet existing approaches suffer from high false positive rates, weak contextual understanding, and limited recall. This work presents the most comprehensive empirical evaluation to date of Small Language Model (SLM) based vulnerability detection, spanning more than twenty system configurations ranging from simple prompting baselines to advanced retrieval-augmented ensembles. Our methodology systematically examines how small LLMs (1–3B parameters) behave under different prompting strategies, static analysis integrations, ensembling techniques, and retrieval mechanisms. Through extensive ablation studies on the DiverseVul dataset, we find that many commonly assumed improvements such as complex Chain-of-Thought prompting, multi-model aggregation, or post-processing verification frequently degrade performance, largely due to overfitting and the inherent reasoning limitations of small models. In contrast, retrieval-augmented generation (RAG) consistently enhances both recall and stability, providing an average F1 gain of 4–5% by grounding model predictions in relevant vulnerability examples. Our best-performing system, a RAG-enhanced stacking ensemble with structured reasoning, achieves a 71.2% F1-score, surpassing the state-of-the-art LLaMA 3.2 model by 5.2 percentage points. our findings demonstrate that effective vulnerability detection does not arise from increased architectural complexity but from combining lightweight models with robust retrieval and context-aware reasoning. This study establishes RAG-driven small LLMs as a practical, cost-efficient, and empirically validated direction for secure software analysis. our source code at https://github.com/rafi79/RAG-Enhanced-Multi-Model-Ensemble-for-Automated-Vulnerability-Detection-Using-SLMs.

The integration of renewable energy sources into microgrids demands advanced power electronic solutions to maintain power quality and stability amidst variable generation and dynamic load conditions. This study focuses on improving a five-level cascaded H-bridge (CHB) inverter to suppress harmonics and regulate voltage in renewable-integrated microgrids. By using a hybrid control strategy that integrates variable frequency pulse width modulation (VFPWM) with proportional-integral (PI) controllers (Kp = 7, Ki = −4), systematically tuned via the Ziegler–Nichols method. An LCL filter (L1 = 16.8 mH, L2 = 5.05 mH, C = 104.9 μF) attenuates harmonics. Performance is validated in MATLAB/Simulink under resistive/nonlinear loads, carrier frequencies (900–1200 Hz), and dynamic microgrid conditions. The significance lies in achieving IEEE 519-compliant 2.64% THDI (under standardized conditions: 10 Ω resistive load, 100 V per module, 1050 Hz carrier, VFPWM), facilitating 40% smaller passive filters and scalable renewable integration for tropical climates like Malaysia. The system achieves 2.64% grid current THDI (10 Ω resistive load, 100 V per H-bridge module, 1050 Hz carrier frequency, VFPWM modulation), reducing total harmonic distortion (THD) by 50%–70% vs conventional three-level inverters. The combination of VFPWM and phase disposition yields optimal voltage THDU (26.50%–26.75%). Optimized PI tuning lowers grid current THDI from 12.54% to 2.64%, complying with IEEE 519. The five-level CHB enhances scalability, fault tolerance, and power quality, enabling cost-effective integration, making it a preferable choice for renewable integration.