UCSC Digital Library Collection:
https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/4574
2023-09-27T02:28:09ZVoice Recognition for User Authentication at Online Examinations
https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/4647
Title: Voice Recognition for User Authentication at Online Examinations
Authors: Wijithapala, W.D.C.P.K.
Abstract: Digitization has made a huge impact on education sector and many institutes all over the world had already transitioned to online from traditional face-to-face lectures. While learning is being practiced online, most of the examinations were conducted in general approaches within the institute’s premises. But the sudden outbreak of COVID-19 pandemic forced all the educators to adapt online platforms within a short period, who were being reluctant to shift earlier. Therefore, the necessity of digital assessment platforms with secure testing environments have arisen not only for educational sector but also in recruitment procedure of employees.
In implementing an online examination system, the prime challenge is to maintain the credibility and the transparency with participants authentication. Therefore, introducing a better approach of user authentication is crucial in every examination platform. Considering the economical and the educational background in Sri Lanka, we are unable to expect the students to have accessibility of specific instruments and high-speed internet. Therefore, this study introduces a voice-based user authentication approach for online examinations which can be acquired with limited facilities.
Voice based user authentication is to identify a user by analyzing the unique features of his /her voice sample. Recent studies show that user authentication with GMM (Gaussian Mixture Models) have efficiently used in speaker recognition. The key features of text independent voice signals are obtained using MFCC (Mel Frequency Cepstral Coefficients) and a unique model for the all the speakers who enrolled to the system is generated using Gaussian Mixture Models. The maximum likelihood algorithms are used to match the users voice samples against the speakers who are already enrolled.
The dataset for the study is obtained from the UCSC/LTRL Speech corpus which contains 60 users with speech utterances of Sinhala language. A Web based application has been developed to implement the user authentication approach using python http (Hypertext transfer protocol) service and PHP. The accuracy of over 90% on correct identifications is obtained by models with voice samples relatively higher duration with GMM and MFCC where 20 trials are tested against the whole set of 60 trained models.2022-08-26T00:00:00ZImpact of Events on Stock Market and Prediction of Impact status and Period
https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/4646
Title: Impact of Events on Stock Market and Prediction of Impact status and Period
Authors: Wijayaweera, V.P.
Abstract: Stock market is a public place where issuing, buying, and selling of a fractional ownership in a company and it is one of the main focuses which could impact for the economy of a country. Key method where investors could profit from purchasing stocks is by selling their stock for a higher price than the purchased price where they can profit if the stock price increases from their initial purchased price. It is harder to predict how the stock rates change time to time and it is certainly a challenge to predict the rates of the shares. When a special event or an incident is happening or happened in the country or in the company, it will affect for the companies which depend or has an impact from the event. Stock prices of a company reacts according to the new information announced by companies or any event which has an impact for the company happens in the country. Usually, it happens after the reaction of the share prices which happened after receiving the new information about an event as investors pay attention to this information so that this will affect for the asset prices dynamically. Hence, the stock rates will be changed accordingly. At times, it will affect for the stock rates drastically after a certain time. The period can be within a day, a week, a month similarly. After an event or an incident occurred, at present there’s no method to predict the bearish or bullish period so that investors couldn’t gain a better or a maximum profit from the stock exchanges. Due to that investors could not get the opportunity to have the maximum profit from the stock trading. If an investor has a doubt on the time for the stock trade, then the maximum profit or rather better profit could not be reached if the stocks are bought or sold within an incorrect time frame.
It would help to achieve more profit for an investor if there’s a way to predict the periods after a certain event where investors can purchase or sell the stocks. Therefore, to cater that problem, in this research project, it will be examined how the events or incidents impact on the market prices which outcome the bearish and bullish periods. According to that examined results, main goal is to predict the bearish and bullish status and the period after an event which impacts to the share prices. After that the results will be evaluated with the actual values to determine the accuracy of the predictions.
It is always a challenge on determining the movement of stock market as it highly depends on traders' emotions, decisions etc. According to the results from this research study, impact and its period did not always outcome with accurate predictions, hence it also shows that sometimes the predictions could be correct and sometimes it could be incorrect.2022-08-26T00:00:00ZIdentification of sticker printing defects in glove manufacturing process using Computer Vision techniques
https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/4645
Title: Identification of sticker printing defects in glove manufacturing process using Computer Vision techniques
Authors: Wickramathunga, W. M. C.
Abstract: This research aims to provide an automatic and real-time defect detection framework for the glove manufacturing industry using computer vision techniques. This research concerns detecting defects on specific sticker printed on the glove at the end of the production. The whole approach is divided into two operational modes: Teaching mode and Inspection mode. The teaching mode contains time complex tasks that can be performed before the actual inspection. The inspection mode does the actual inspection to find the defects.
An image of a printed sticker will be processed in inspection mode using three levels to identify defects. Lower levels contain naïve computer vision algorithms and detect high-degree errors only, whereas higher levels contain complex algorithms that could detect more sophisticated errors. It is an efficient technique to identify defects in the early stages of the defect inspection process.
The significance of sticker's content to its domain will be calculated for every object in the sticker by combining the visibility and domain importance of that specific content. The visibility of content is measured using size and density. A decision function is proposed to decide whether to accept or reject the glove by considering the calculated error and the significance. Finally, a quality measurement model is proposed to calculate the printed sticker's quality for each accepting glove.
The visibility calculation model proved to be valid and consistent with perceptual visibility. The significance calculation model also provides reliable and consistent results according to testing. The defect inspection process is also efficient and performs as expected. However, inspection level-3 provide inconsistent result in some situations, and that algorithm needs to be improved.2022-08-26T00:00:00ZPre-Detection of Dementia Using Machine Learning Mechanism
https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/4644
Title: Pre-Detection of Dementia Using Machine Learning Mechanism
Authors: Weerasinghe, M.S.U
Abstract: Dementia is a neurological disorder that affects millions of people worldwide. Dementia is also not a normal part of aging, and there is no cure or effective treatment for it. The number of persons suffering from dementia is rapidly increasing. According to the 2015 World Alzheimer Report, there are 46.8 million people worldwide who have been diagnosed with dementia, with that figure anticipated to rise to 74.7 million by 2030 and 131.5 million by 2050. The number of people diagnosed with dementia in Sri Lanka is continuously increasing. According to government projections, there are already more than 0.2 million dementia patients, with that number anticipated to climb to 0.5 million by 2050. As a result, dementia has emerged as a serious medical condition that needs to be addressed for the sake of society's well-being.
Dementia is a difficult condition to manage, and it must be dealt with quickly. According to the World Alzheimer’s Association, Dementia is one of the most financially costly diseases in the world since there is no proper treatment to cure the disease and the cause of the disease is not identified correctly.
Recently there is a strong interest in machine learning mechanism which provides a better classification accuracy than the conventional classification methods. Based upon the recent studies we design and perform some experiments to investigate the possibility of early diagnosis of dementia from machine learning mechanism using the clinical data.
In this project, we compare machine learning algorithms to clinical data from dementia patients in order to develop a better approach for detecting dementia at an early stage. We primarily use three main advanced machine learning algorithms: SVM, decision tree, and random forest.
We train the preprocessed clinical data with the above advance algorithms and takes the highest accurate algorithm after comparing the results along with the confusion matrix of each other. After comparing the confusion matrix results, we choose random forest algorithm as the most accurate machine learning algorithm and it used to trained machine learning model.
In this project we developed a mobile application for the public people and this mobile application is integrated with the machine learning model. So it is very helpful for the general public to identify their day-to-day mental capabilities with this mobile application. This will evaluate the person's dementia level and aid in the early detection of dementia. As a result, this research will be a significant step forward in the prevention of dementia in Sri Lanka.
We expect that this dissertation will help researchers to get better understand about how machine learning can be used in early dementia detection.2022-08-26T00:00:00Z