Search

Deep learning (DL) networks have grown into powerful alternatives for speech enhancement and have achieved excellent results by improving speech quality, intelligibility, and background noise suppression. Due to high computational load, most of the DL models for speech enhancement are difficult to implement for realtime processing. It is challenging to formulate resource efficient and compact networks. In order to address this problem, we propose a resource efficient convolutional recurrent network to learn the complex ratio mask for real-time speech enhancement. Convolutional encoder-decoder and gated recurrent units (GRUs) are integrated into the Convolutional recurrent network architecture, thereby formulating a causal system appropriate for real-time speech processing. Parallel GRU grouping and efficient skipped connection techniques are engaged to achieve a compact network. In the proposed network, the causal encoder-decoder is composed of five convolutional (Conv2D) and deconvolutional (Deconv2D) layers. Leaky linear rectified unit (ReLU) is applied to all layers apart from the output layer where softplus activation to confine the network output to positive is utilized. Furthermore, batch normalization is adopted after every convolution (or deconvolution) and prior to activation. In the proposed network, different noise types and speakers can be used in training and testing. With the LibriSpeech dataset, the experiments show that the proposed real-time approach leads to improved objective perceptual quality and intelligibility with much fewer trainable parameters than existing LSTM and GRU models. The proposed model obtained an average of 83.53% STOI scores and 2.52 PESQ scores, respectively. The quality and intelligibility are improved by 31.61% and 17.18% respectively over noisy speech.

Background: Upper urinary tract obstruction is a serious condition that can lead to significant renal impairment if not managed promptly. Emergency urinary diversion is crucial for alleviating obstruction and preserving renal function. This study investigates various emergency urinary diversion techniques, compares their outcomes, and evaluates factors influencing recovery. Methods: We conducted a prospective study involving 106 patients diagnosed with obstructive uropathy at Jawaharlal Nehru Medical College, Aligarh Muslim University. Patients were classified into upper urinary tract obstruction and lower urinary tract obstruction. 70 patients of the upper urinary tract obstruction cases were further managed by either percutaneous nephrostomy (PCN) or double-J (DJ) ureteral stenting diversion procedures. 68 out of 70 upper urinary tract obstruction underwent either of the two procedures. Clinical, demographic, and laboratory parameters were analysed in such patients, and their outcomes were assessed at one and three months. Results: The study included 68 out of 106 patients with a mean age of 48 years. Most patients presented with acute kidney injury (AKI) and had symptoms of loin pain and haematuria. Both PCN and DJ stenting were equally effective in managing upper tract obstruction, with no significant differences in improvement of the laboratory parameters (e.g. Hemoglobin, Serum Creatinine, estimated Glomerular Filtration Rate, Blood Urea Nitrogen) between the two techniques at one and three months. Associated factor like age was associated with unfavourable outcomes. Conclusion: Emergency urinary diversion via PCN or DJ stenting effectively manages upper urinary tract obstruction, with comparable outcomes. Early intervention and management of underlying conditions are critical for optimal renal recovery and prevent irreversible kidney damage.

The purpose of this study is to explore how internal and external dynamic capabilities and green human resource (HR) practices enhance sustainable outcomes in SMEs. Data were collected using a survey methodology targeting employees of SMEs located in the northern region of Bangladesh and West Bengal, India. A total of 327 employees participated in the survey conducted between May and June 2024. A five-point Likert scale was employed to assess the responses, and regression analysis was used to test the study’s hypotheses. The analysis was conducted at a 5% significance level, and the data were analyzed using the SPSS software package to ensure accuracy and reliability. The results indicate that internal and external dynamic capabilities, as well as green HR practices, play a significant role in shaping the sustainable performance of SMEs. External dynamic capabilities (β = 0.498, p < 0.05**) have the highest effect on the sustainable performance of SMEs, and internal capability (β = 0.384, p < 0.05**) has the second highest effect on SME performance. Additionally, a significant positive correlation (r = 0.498) was found between green HR practices and the overall performance of SMEs. This study contributes to the growing body of literature by providing empirical evidence on the critical factors driving sustainable performance in SMEs. It highlights the importance of integrating dynamic capabilities and environmentally conscious HR practices to achieve long-term sustainability in the SME sector, particularly in the context of developing countries like Bangladesh and India.

Traditional recurrent neural networks (RNNs) encounter difficulty in capturing long-term temporal dependencies. However, lightweight recurrent models for speech enhancement are important to improve noisy speech, while being computationally efficient and able to capture long-term temporal dependencies efficiently. This study proposes a lightweight hourglass-shaped model for speech enhancement (SE) and automatic speech recognition (ASR). Simple recurrent units (SRU) with skip connections are implemented where attention gates are added to the skip connections, highlighting the important features and spectral regions. The model operates without relying on future information that is well-suited for real-time processing. Combined acoustic features and two training objectives are estimated. Experimental evaluations using the short time speech intelligibility (STOI), perceptual evaluation of speech quality (PESQ), and word error rates (WERs) indicate better intelligibility, perceptual quality, and word recognition rates. The composite measures further confirm the performance of residual noise and speech distortion. With the TIMIT database, the proposed model improves the STOI and PESQ by 16.21% and 0.69 (31.1%) whereas with the LibriSpeech database, the model improves STOI by 16.41% and PESQ by 0.71 (32.9%) over the noisy speech. Further, our model outperforms other deep neural networks (DNNs) in seen and unseen conditions. The ASR performance is measured using the Kaldi toolkit and achieves 15.13% WERs in noisy backgrounds.

Computer vision is witnessing a surge of interest in machines accurately recognizing and interpreting human emotions through facial expression analysis. However, variations in image properties such as brightness, contrast, and resolution make it harder for models to predict the underlying emotion accurately. Utilizing a robust architecture of a convolutional neural network (CNN), we designed an efficacious framework for facial emotion recognition that predicts emotions and assigns corresponding probabilities to each fundamental human emotion. Each image is processed with various pre-processing steps before inputting it to the CNN to enhance the visibility and clarity of facial features, enabling the CNN to learn more effectively from the data. As CNNs entail a large amount of data for training, we used a data augmentation technique that helps to enhance the model's generalization capabilities, enabling it to effectively handle previously unseen data. To train the model, we joined the datasets, namely JAFFE and KDEF. We allocated 90% of the data for training, reserving the remaining 10% for testing purposes. The results of the CCN framework demonstrated a peak accuracy of 78.1%, which was achieved with the joint dataset. This accuracy indicated the model's capability to recognize facial emotions with a promising level of performance. Additionally, we developed an application with a graphical user interface for real-time facial emotion classification. This application allows users to classify emotions from still images and live video feeds, making it practical and user-friendly. The real-time application further demonstrates the system's practicality and potential for various real-world applications involving facial emotion analysis.

IntroductionAlzheimer's disease (AD) is a neurodegenerative disorder and the most prevailing cause of dementia. AD critically disturbs the daily routine, which usually needs to be detected at its early stage. Unfortunately, AD detection using magnetic resonance imaging is challenging because of the subtle physiological variations between normal and AD patients visible on magnetic resonance imaging.MethodsTo cope with this challenge, we propose a deep convolutional generative adversarial network (DeepCGAN) for detecting early-stage AD in this article. The DeepCGAN is an unsupervised generative model that expands the dataset size in addition to its diversity by utilizing the generative adversarial network (GAN). The Generator of GAN follows the encoder-decoder framework and takes cognitive data as inputs, whereas the Discriminator follows a structure similar to the Generator's encoder. The last dense layer uses a softmax classifier to detect the labels indicating the AD.ResultsThe proposed model attains an accuracy rate of 97.32%, significantly surpassing recent state-of-the-art models' performance, including Adaptive Voting, ResNet, AlexNet, GoogleNet, Deep Neural Networks, and Support Vector Machines.DiscussionThe DeepCGAN significantly improves early AD detection accuracy and robustness by enhancing the dataset diversity and leveraging advanced GAN techniques, leading to better generalization and higher performance in comparison to traditional and contemporary methods. These results demonstrate the ecacy of DeepCGAN in enhancing early AD detection, thereby potentially improving patient outcomes through timely intervention.

This empirical study aims to explore the factors that impact job satisfaction among employees in private banks in Bangladesh, considering the prevailing atmosphere of intense competition in the country’s banking sector. Recognizing that employees are a crucial asset to any organization’s success, it is essential to explore the variables that contribute to their work contentment. The methodology adopted for this study follows a quantitative approach, employing regression analysis for scrutinizing the relationships between job satisfaction and selected variables. This study included a sample of 320 individuals employed in private banks in Bangladesh. This sample size was considered suitable for conducting a robust quantitative regression analysis. To collect the necessary data, the questionnaire was given to the bank personnel and employees via their email addresses. The survey instrument utilized in this investigation was modified from earlier research endeavors to ensure validity and reliability. To analyze the collected data, the researchers utilized the statistical software SPSS version 29.0. The regression analysis feature of SPSS was employed to test the model and hypotheses formulated for this study. The findings of the study reveal that green HR practices (β = 0.284), which encompass environmentally friendly policies and practices in human resource management, significantly contribute to employee job satisfaction. Additionally, remuneration (β = 0.224) emerges as another vital factor influencing job satisfaction. Furthermore, work environment (β = 0.298) was found to have a significant positive and highest effect among other two variables on job satisfaction levels.

This study aimed to delineate groundwater recharge zones using a combination of analytical hierarchy process (AHP), fuzzy-AHP, and frequency ratio (FR) models. Additionally, it aimed to compare the effectiveness of these models in groundwater recharge potential zone mapping. To achieve these objectives, nine groundwater influencing factors were considered, including geology, soil types, lineament density, elevation, slope, topographic wetness index, drainage density, land use land cover, and rainfall. Thematic maps for all these factors were generated using satellite and conventional data in the ArcGIS environment. Weight was assigned to each thematic layer based on its significance to recharge. All thematic layers were combined using AHP model-l (WLC), AHP model-ll (Weighted sum), fuzzy-AHP overlay, and FR-based model using ArcGIS. The findings revealed that 15% and 39% of the study area have high recharge potentials according to AHP-based model-l and model-ll, respectively. The FAHP model demarcated 43% of the area as high recharge zones while the FR model demarcated 42% of the area as high recharge zones. The majority of high groundwater recharge areas were found in the central part of the study area, while the southern part was demarcated as a moderate recharge zone. The eastern and western parts were demarcated as low recharge potentials zones. To validate the accuracy of these models, the study used receiver operating characteristic (ROC) validation curves. The ROC curves revealed that AHP model-ll had the highest accuracy (AUC=89%) followed by the FAHP model (AUC=88%), AHP model-l (AUC=84%), and FR (AUC=81%)...

The increasing trend in sustainable economic growth over the last few decades has elevated the energy demand, technological innovation, and access to minerals resources are contributing well to economic development. This article investigates the nexus among minerals resource complexity, energy consumption, technology, and economic growth by employing autoregressive distributed lag and vector error correction techniques for Pakistan from 1995 to 2021. Following thorough research, the long-term results show that an important 9.73 points of economic growth result from every 1 % increase in the complexity of natural resources. On the other hand, technology and energy use negatively affect economic growth, causing drops of −0.03 and −12.9 points, respectively. One-way causality was noted between mineral resources' complexity and economic growth. Moreover, a one-sided causality effect was also confirmed between energy use, technology, and economic growth. Additionally, it was predicted that there is a neutral causality between mineral resources and technology. Corresponding to this, technology and energy consumption have a bidirectional causal relationship. These results imply that energy consumption, technological advancements, and mineral resources contribute as major economic growth drivers and can improve environmental quality.

Introduction Patient satisfaction is an integral part of healthcare quality, impacting treatment adherence, patient loyalty and healthcare utilisation. Parental satisfaction is particularly crucial as parents influence decision-making for their child’s healthcare.Objective The objective was to assess parental satisfaction levels (as measured by Patient Satisfaction Questionnaire-18 (PSQ-18)) related to paediatric healthcare services and identify specific factors significantly influencing parental satisfaction with paediatric healthcare services.Methodology This is a descriptive cross-sectional study that used a structured questionnaire based on the PSQ-18. The study was conducted in Faisalabad, Islamabad, Peshawar and Swat. Seven subscales measured satisfaction across various dimensions.Results We found 882 suitable responses indicating a diverse participant demographic, with the largest group in the 6–12 years age category. The overall mean parental satisfaction score was 2.0±0.5 (40.0%). Notably, financial aspects scored highest at 2.8 (55%), followed by accessibility and convenience at 2.0 (40.5%). However, lower satisfaction scores in the interpersonal relationship (1.8±0.5) and technical quality (1.8±0.5) domains were recorded. Sociodemographic analysis indicated age, education and occupation significantly influenced satisfaction. Conversely, factors such as gender, residence and the presence of chronic disease did not significantly impact satisfaction levels.Conclusion The study offers valuable insights into paediatric patient satisfaction in Pakistan, emphasising the need for targeted interventions and improvements in specific domains to enhance overall healthcare quality.

Abstract Climate change policy has several potential risks. The purpose of this study is to investigate the impact of green technology development, green energy consumption, energy efficiency, foreign direct investment, economic growth, and trade (imports and exports) on greenhouse gas (GHG) emissions in South Asia from 1981 to 2018. We employed Breusch Pagan LM, bias-corrected scaled LM, and Pesaran CD as part of a series of techniques that can assist in resolving the problem of cross-sectional dependence. First and second generation unit root tests are used to assess the stationarity of the series, Pedroni and Kao tests are used to test co-integration. The long-term associations are examined using fully modified ordinary least square (FMOLS) and panel dynamic ordinary least square (DOLS) for robustness. The results revealed that trade, growth rate, and exports significantly increase GHG emissions. This accepted the leakage phenomenon. The results also demonstrated that green technology development, green energy consumption, energy efficiency, and imports all have a significant negative correlation with GHG emissions. Imports, advanced technical processes, a transition from non-green energy to green energy consumption, and energy efficiency are thus critical components in executing climate change legislation. These findings highlight the profound importance of green technology development and green energy for ecologically sustainable development in the South Asian countries and act as a crucial resource for other nations throughout the world when it comes to ecological security. This research recommends the consumption of environmentally friendly and energy-efficient technologies in order to mitigate climate change and the government's implementation of the most recent policies to neutralize GHG emissions in order to achieve sustainable development.

PurposeThe goal of this study is to investigate treatment planning of total marrow irradiation (TMI) using intensity-modulated spot-scanning proton therapy (IMPT). The dosimetric parameters of the intensity-modulated proton plans were evaluated and compared with the corresponding TMI plans generated with volumetric modulated arc therapy (VMAT) using photon beams.MethodsIntensity-modulated proton plans for TMI were created using the Monte Carlo dose-calculation algorithm in the Raystation 11A treatment planning system with spot-scanning proton beams from the MEVION S250i Hyperscan system. Treatment plans were generated with four isocenters placed along the longitudinal direction, each with a set of five beams for a total of 20 beams. VMAT-TMI plans were generated with the Eclipse-V15 analytical anisotropic algorithm (AAA) using a Varian Trilogy machine. Three planning target volumes (PTVs) for the bones, ribs, and spleen were covered by 12 Gy. The dose conformity index, D80, D50, and D10, for PTVs and organs at risk (OARs) for the IMPT plans were quantified and compared with the corresponding VMAT plans.ResultsThe mean dose for most of the OARs was reduced substantially (5% and more) in the IMPT plans for TMI in comparison with VMAT plans except for the esophagus and thyroid, which experienced an increase in dose. This dose reduction is due to the fast dose falloff of the distal Bragg peak in the proton plans. The conformity index was found to be similar (0.78 vs 0.75) for the photon and proton plans. IMPT plans provided superior superficial dose coverage for the skull and ribs in comparison with VMAT because of increased entrance dose deposition by the proton beams.ConclusionTreatment plans for TMI generated with IMPT were superior to VMAT plans mainly due to a large reduction in the OAR dose. Although the current IMPT-TMI technique is not clinically practical due to the long overall treatment time, this study presents an enticing alternative to conventional TMI with photons by providing superior dose coverage of the targets, increased sparing of the OARs, and enhanced radiobiological effects associated with proton therapy.