2015

Browse

Filters

20153

Settings

Search Results

Now showing 1 - 3 of 3
  • Thumbnail Image
    Item
    Enhancing Software Development Process (ESDP) using Data Mining Integrated Environment
    (Department of Computer Science and Engineering (CSE), Islamic University of Technology (IUT), Board Bazar, Gazipur-1704, Bangladesh, 2015-11-15) Rahman, Ziaur
    Nowadays, it has become a basic need to reuse existing Application Pro- gramming Interface (API), Class Libraries and frameworks for the rapid software development. Software developers often reuse this by calling the respective APIs or libraries. But in doing so, developers usually encounter di erent problems in searching for appropriate code snippets. In most cases API and Libraries are com- plex and not well structured. Online search engine consumes time in searching, yet match is not that relevant and representation is not good. To get a suggestion according to the query we can nd that snippet using code search engines. In some cases database dependent searching and remote web server based mined repository searching bring problem to the developers. Finding an API recommendation in code search engine often deal with extra-large les that eventually slows down soft- ware development process. We have searched for a solution throughout our work and tried to bring a better outcome. As an alternative action we have implemented a system what we call \Enhancing Software Development Process (ESDP)" tool that is able to provide an e cient and working integrated environment to the developers with a better abstraction and representation of the search results and programmer's need to be derived from the source codes. We also have built and applied an XML based enriched repository to get recommendation from the mined repository in the client side without interacting with the Internet dependent server to save complications and times. ESDP provides the most relevant code skeletons or mapping to developers using graph based representation. We have evaluated that ESDP boosts up the software development process enough particularly by reducing the response time in the coding phase. By giving a number of queries for an API, ESDP gathers more relevant source snippets through data mining. We have evaluated the e ciency of ESDP tool using a set of various queries and com- pared with the other existing tools. The results show that in di erent experiments ESDP consumes quite less time than some of the updated approaches to nd the code snippet solution.
  • Thumbnail Image
    Item
    Receiver Assisted Rate Adaptation in Wireless Networks
    (2015-11-15) Rashid, Nafiul
    The IEEE 802.11 wireless local area network (WLAN) standard, especially 802.11a re- mains the most popular way to exchange data over wireless links. The major requirement is to adapt to highly dynamic channel conditions with minimum overhead and ensure robustness and speed of transmission. To this end, we propose a novel rate adaption scheme NARC (Neighbor Aware Rate Control). Firstly, our key contributions include exploiting the more precise channel estimation of SNR based rate adaptation coupled with estimating the channel condition at the receiver and nally sending this estimated information to the transmitter with minimum overhead. We use acknowledgment rates to serve this feedback purpose. Our feedback mechanism also allows for optimal rate switch rather than sequential one that most of the existing methods support. Secondly, we address the stale feedback problem that the SNR based methods mainly su er from and provide a unique solution to overcome this. To the best of our knowledge no works have addressed the solution to this problem. The stale feedback problem was mitigated by a prediction mechanism using linear regression on the observed rates on sender side and feedback rates from the receiver side. Besides, we di erentiate the cause of frame loss as either due to channel error or collision using RTS/CTS but in an adaptive fashion to minimize overhead but at the same time ensure that rate is not falsely changed due to frame loss caused by collision. NARC exploits the best of SNR based approaches and provides channel condition at the receiver to the transmitter with minimum overhead thereby ensuring optimal rate switching decision aided by sender side prediction mecha- nism to tackle against stale feedback problem. Moreover use of Adaptive RTS provides robustness to our method.
  • Thumbnail Image
    Item
    Optical Flow Based Facial Expression Recognition from Video Sequences
    (Department of Computer Science and Engineering (CSE) Islamic University of Technology (IUT), Board Bazar, Gazipur-1704, Bangladesh, 2015-11-15) Salekin, Md Sirajus
    Facial expression is one of the most powerful masses of non-verbal communication through which we can easily enter the world of one's instant emotions or intuitions. As most of the time, it elicits naturally, so it brings out a lot of applications in the eld of machine intelligence, behavioral science, clinical practices, biometric security, gaming, human computer interactions, psychological research, data-driven animation etc. Proper expression recognition can lead to an intelligent machine for taking commands more e ectively, can show the insight psychological condition of a patient to a psychiatrist or researcher, can show the next suggested path for any computer game. But automatic facial expression recognition is a challenging task due to the di erent factors such as variations in illumination, pose, facial expression, alignment, di erent ages, occlusions etc. In this thesis paper, we propose a novel feature representation by a new feature descriptor, named Patterns of Oriented Motion Flow (POMF) from the optical ow information, to recognize the proper facial expression from a facial video. The POMF computes di erent directional motion information and encodes those directional ow information with enhanced local texture micro pattern. As it captures the spatial temporal changes of facial movements through optical ow and enables to observe both local and global structures, it shows its robustness for the facial expression. Finally, the POMF histogram is used to train the expression model by Hidden Markov Model (HMM). To train through the HMM, the objective sequences are produced by the generation of codebook using K-means clustering technique. The performance of the proposed method has been evaluated over the RGB camera based and Depth camera based video. We also compare the proposed method with the other promising appearance based methods. Experimental results demonstrate that the proposed POMF descriptor is more robust in extracting facial information and provides higher classi cation rate compared to other existing promising methods.