Browsing by Author "Rahman, Md. Zahidur"

Now showing 1 - 5 of 5

An alternate to hawk eye using the graphics processing unit
(BRAC University, 4/28/2013) Ehsan, Radin Ahmed; Rahman, Md. Zahidur
The more powerful and advanced our processing units become our hunger for more computation power increases. While we are pushing the Central Processing Unit (CPU) to its last limit in the hope to get more computational power, the Graphical Processing Unit (GPU) is sitting idle most of the times. The GPU is immensely powerful in terms computation power and as most modern day GPU’s have multi-core architecture; parallel computing can be performed easily. This paper will look into the prospect of improving Hawk Eye, a technology currently used as decision aid system in cricket and tennis in terms of accuracy. For the system NVDIA GPU has been used on CUDA platform.
Effect of different softeners on moisture absorption and transmission properties of knitted cotton fabric
(Daffodil International University, 2016-10-26) Roy, Shibashis; Rahman, Md. Zahidur
The influence of softener on different moisture properties were experimented in this project work. Three types of softener (Nonionic, Cationic and Polyethylene emulsion) were selected for this purpose. Results showed that water vapor transmission rate increases with time for all types of softener application but the rate was far lower for the polyethylene emulsion than the others. Moreover, transmission rate increases with the concentration of nonionic softener but decreases in cases of cationic softener. On the other hand, cationic softener showed low moisture regain and moisture content as well as a lower wicking property which influences its lower drying time than the others.
Human Activity Recognition using wearable body sensor by machine learning approach
(BRAC University, 2019-12) Promi, Sadia Tangim; Rahman, Md. Zahidur; Mostafa, Moumita; Harun, Sarah Bintay; Alam, Md. Golam Rabiul
The prevalence of electronics devices and the increase in computer resources, like networking, storage, accessibility and sensor capacity, have significantly improved the lives of humans. Now a days most smart devices have a number of strong sensing equipment, such as sensors for movement, position, connection and direction.Basically, movement or motion tracking sensors are commonly been using to classify the physical activities of humans. This has opened entryways for a wide range of and intriguing applications with regards to a numerous zones, for example, human healthcare well being and transportation, security system. In this point of view, this research gives a complete, best in class audit of the present circumstance of human activity recognition (HAR) approaches with regards to inertial sensors in electronic portable smartphone devices. Our research started by analyzing the principles of human activities and the entire historical events based on electronics deices such a smartphone, which demonstrate the development in this area over the past few years. Our approach concentrates on the introduction of the means of HAR arrangements with regards to sensors. We propose a methodology which incorporates traditional signal processing techniques with deep learning tools to robustly classify activities from wearable body sensor data. Our proposed methodology achieves a validation accuracy of 96.26% in the WISDM Dataset and is able to recognize human activity from wearable body sensor data robustly.
Improving product rating system with text mining on product review/comments
(BRAC University, 4/30/2014) Karim, Md. Ahsanul; Islam, Raduanul; Ahmed, Sk. Wasif; Rahman, Md. Zahidur
It is often considered a better exercise to have a complete idea of a particular service or a product before availing it. Now a days almost every online shopping sites or even the manufacturer of the product has a star based rating system and review/comment zone in their website. It is often not feasible to go through all the review before purchasing or availing that particular product. So people often tends to have an idea based on the number of stars on that product.Currently available systems use a star based rating where people rate the service or product on the scale of 5 or 10. The problem with that is when they give those rating stars they often tends to give it without giving much thought to it. User experience level and his mind set while rating varies very much. For example a person who loves particular brand of Soda, if he drink soda of another brand he might rate it lower than what it should be because he is used to a particular brand. But when he writes a review the chances are higher that he will write the major positive and negative aspects of that product . Thus the chances of getting a better feedback comes when it is review rather than stars. But as it has been mentioned earlier, going through all the reviews are not feasible. So we tried to improve the rating system by extracting information from the review text by using text mining technique upon that.
Parallel computing using GPU for efficient traffic simulation
(BRAC University, 2013) Ahmed, Sadat Sakif; Rahman, Md. Zahidur
Parallel Computing can be made possible using the multiple cores of the Graphics Processing Unit (GPU) thanks to the modern programmable GPU models. This allows the use of parallel computing techniques to improve upon the computation time of large scale traffic simulations. This paper proposes the use of a multi-processor algorithm for creating efficient traffic simulation software. The method in consideration achieves this by separating the road network into regions which are individually computed as a threaded block inside the GPU and merged together using the Central Processing Unit to provide the final data of the simulation. A significant improvement in the computation time is observed when the proposed parallelization techniques are applied to the simulator.