International Islamic University, Chittagong

About: International Islamic University, Chittagong is a education organization based out in Chittagong, Bangladesh. It is known for research contribution in the topics: Debye model & Density functional theory. The organization has 1111 authors who have published 1089 publications receiving 7918 citations. The organization is also known as: IIUC.

Variation of structural and mechanical properties as 20 wt% ZrO 2 added to Al 2 O 3 for biomedical application

Abstract: Employing slurry method, 20 wt % of ZrO 2 has been prepared by incorporating high purity nanocrystalline a-alumina powder and sintered at 1600°C for two hours. XRD pattern confirmed the existence of m and t phase of 20 wt % of ZrO 2 compared to pure alumina. The theoretical and bulk density increases with the doping of zirconia, whereas porosity declines with the similar change of zirconia. By using SEM, the surface morphology of the samples has been studied. The effect of zirconia content on hardness and elastic modulus were examined. Maximum hardness and elastic modulus has been observed for pure alumina. Moreover, flexural strength shown increasing trend with the rise of zirconia content. An alternative design of Al 2 O 3 -ZrO 2 composites can be employed for improved structural and mechanical properties for biomedical application rather than using pure alumina at 1600°C.

Recognition of Pedestrian Clothing Attributes from Far View Images using Convolutional Neural Network

Abstract: Attributes of pedestrians is a very important factor when it comes to the task of re-identifying pedestrians or distinguishing one pedestrian from another. Different types of clothing on a pedestrian are more recognizable than other attributes. Recognizing multiple attributes together explores the relationship among them and their contribution towards each other's representation. We address the pedestrian attribute recognition from far view images task as a multi label single class classification problem and approach it with a specialized convolutional neural network model that achieves an overall accuracy of 90.16% for 34 pedestrian clothing attributes including hair as defined in the PETA dataset which consists of far-view images. We compare the recognition accuracy of the attributes that were previously worked with in some of the previous CNN models. Our model outperforms these previous models in the overall accuracy and for more than half of the attributes.

Measuring the Factors Affecting Employees Job Satisfaction in Readymade Garments Industry: Bangladesh Perspective

Abstract: : This paper attempts to identify the factors influencing employee’s job satisfaction of readymade garment sector (RMG) in Bangladesh. To lookout this point, hundred workers employed in RMG sector in Chittagong were selected by random sampling technique. The study is based on primary data which has been collected through structured questionnaire. The gathered data were assessed by using statistical software (SPSS). Findings of the study indicate that job satisfaction in RMG sector is mostly influenced by salary, promotion, housing facility, behavior of supervisor and job security. However workers job satisfaction are less influenced by transport facility and training and development. This study will help the policy makers to make comprehensive planning and strategy to satisfy their employees and thus, it will help for the overall development of this industry.

Evaluation of nonlinear propagation effects on coherent optical transmission systems for various commercial optical fiber types

Abstract: This paper presents the study of Gaussian Noise (GN) model based on power spectral density (PSD) of amplified spontaneous emission (GASE) noise and nonlinear interference (GNLI). GASE is the summation of erbium doped fiber amplifier (EDFA) noise in the optical link. GNLI appears for a number of channels in coherent optical transmission systems. We show the computed nonlinear effects on coherent uncompensated optical transmission systems (CUOTS) for different commercial optical fiber types (PSCF150, SSMF, SSMFULL) considering channel spacing, number of span, span length and number of channels respectively.

Smart Meter with Load Prediction Feature for Residential Customers in Bangladesh

Abstract: Conventional energy meter needs to be monitored physically to make the monthly bill. In substitute of the conventional method, a prepaid meter is installed to calculate the amount of the power consumption and the bill is paid earlier by input a series of digits with the help of meter Dialpad. To know the quantity of the used power, the customer requires to check the energy meter manually through dialing some predefined code. This research paper demonstrates a load forecasting based smart energy meter offering an accurate calculation of the used power and cost, reducing the labor cost, efficient and cost-effective. A microcontroller is engaged in calculating the current, voltage, energy and cost while a GSM module for informing the customers and power distributors about daily power use as well as monthly billing by the SMS system. To illustrate the real-time current, voltage and energy for customer notification a 16×2 Liquid Crystal Display is utilized. An immaculate multimeter that is superbly lined up with the wattmeter is utilized to calculate the current and voltage. To adjust the proposed wattmeter, a variable rheostat has been used where its outcome is nonlinearly contrasted with a conventional wattmeter and obtained the accuracy is about 98%. In addition, forecasting of the following day power consumption makes the proposed framework more exceptional. In consequence of this feature, the system will assist the customer to be economical in power consumption and the amount of power wastage will be diminished drastically. The proposed method is perfect in billing procedure, informing and forecasting the power consumption and enhances the customer satisfaction level.