Chittagong University of Engineering & Technology

About: Chittagong University of Engineering & Technology is a education organization based out in Chittagong, Bangladesh. It is known for research contribution in the topics: Renewable energy & Dielectric. The organization has 1200 authors who have published 1444 publications receiving 10418 citations. The organization is also known as: Engineering College, Chittagong & Bangladesh Institute of Technology, Chittagong.

Effect of growth temperature on the structural and optical properties of CdS:O thin films for CdTe solar cells

Abstract: Oxygenated Cadmium Sulfide (CdS:O) is used as a window layer for high-performance CdTe thin-film solar cells. In this work, the effect of growth temperature on the physical properties of (CdS:O) layer is reported. The CdS:O layer was synthesized on commercially available soda-lime glass (SLG) by using a reactive radio frequency (RF) sputtering in 1.5% oxygen (O 2 )/argon (Ar) ambient as a function of substrate temperature (25 °C to 300 °C). The structural and optical properties of the as-grown CdS:O thin films were carried out by X-ray diffraction and UV-VIS-Spectroscopy respectively. The XRD patterns revealed intriguing structural natures of the deposited films such as amorphous/semi-amorphous nature in the substrate temperature range 25–150 °C, then polycrystalline nature in the range 200–250 °C and again amorphous nature at 300 °C. The transmission spectra exhibited higher optical transmission in the visible wavelength which signifies the enhancement of photo generated current of CdTe solar cells. The bandgap of room temperature (RT) deposited CdS:O thin film for 1.5% O 2 /Ar was found to be increased 12.50% in comparison with CdS film deposited in pure Ar ambient. Besides, the bandgap of CdS:O films for 1.5% O 2 /Ar was decreased with the increase of the substrate temperature. Thus, it has been concluded that the CdS:O thin film deposited at RT found to be better for CdTe solar cell application.

An Empirical Robotic Framework for Initiating Interaction with the Target Human in Multiparty Settings

Abstract: Currently work in robotics is expanding from industrial robots to robots that are employed in the living environment. For robots to be accepted into the real world, they must be capable to behave in such a way that humans do with other humans, and initiate interaction in the same way. This paper focuses on designing a robotic framework that is able to perform an important social function such as, initiating interaction with the target human in multiparty setting in a natural and social ways. However, it is not easy task for the robot to initiate an interaction process with a particular human, especially when the robot and the human are not facing each other or the intended target human is intensely engaged his/her work. In order to initiate an interaction process, we propose a framework that consists of two parts: capturing attention and ensuring attention capture. Evaluation experiments reveal the effectiveness of the proposed system in four viewing situations namely, central field of view (CFOV), near peripheral field of view (NPFOV), far peripheral field of view (FPFOV), and out of field of view (OFOV) respectively.

Classification of Attentional Focus Based on Head Pose in Multi-object Scenario

Abstract: Recently determining of visual focus of attention has gained increased consciousness in computer vision to develop non-verbal communication based system. In this paper, we proposed a computer vision based approach to classify the focus of attention of human in multi-object scenario. In order to determine the current focus of attention head pose is used. To classify the different attentional direction the system is trained supervised machine learning and geometrical analysis techniques. The proposed system is trained with more than 7 h live videos with 9 head poses that contains 435000 frames. The proposed attention classification model achieved 97.00% accuracy on test set with 81000 video frames and visual focus of attention accuracy near to 95.00% with multi-object scenarios.

A Star-Fish Shaped Compact Dual band Antenna for Millimeter Wave Band Applications

Abstract: In this paper a star fish shaped microstrip patch antenna is proposed and its design procedure and characteristics are analyzed. The shape of the antenna patch is similar to a star fish, therefore the antenna is called Star fish antenna. The proposed antenna is able to provide dual band operation in V band (40–52 GHz) and W band (77–95 GHz) which lie in the range of millimeter wave frequency. The proposed antenna has 8 branches of unequal length in order to achieve wide bandwidth in millimeter wave range. Each of the branch resonates at different frequency which ultimately results in a wide bandwidth of 30 GHz with an efficient gain of 7.9026 dBi and highly reduced return loss. Designed antenna has a very compact dimensions of 9×9×2.03 mm3, thus it can be easily fabricated with small handset devices.

Talking vs Non-Talking: A Vision Based Approach to Detect Human Speaking Mode

Abstract: Human talking mode detection is an important issue in human-computer interaction. In this work, we propose a method for detecting human talking and non talking mode detection based on supervised machine learning approach. Visual lip information of human is considered as an important clue. Our goal is to develop a method for human talking and non talking mode detection in real time using supervised classification algorithm. We tested our experiment with a single speaker task and compared the results with the previous method. The results show that our approach can obtain a 98.00% accuracy and a fast executed time.