Browsing by Author "Kaushalya, T"

Now showing 1 - 3 of 3

item: Conference-Extended-Abstract
A system for automatic music transcription
(2011) Abeykoon, H; Kaushalya, T; Akram, N; Weerawarana, S; De Silva, C; Dissanayaka, A
Music notations can be considered as very important information for musicians as well as for music fans, To recreate music which WOJ heard before, one has to know the musical notes which were included in that music recording. For many years computer scientists and engineers have tried to come up with various techniques to automate the task of finding out musical notes from a music recording. Many digital formats which facilitate storing and encoding of music information exist. Many statistical methods have been proposed in literature. But implementation specific detail is very scarce. With this paper we try to address that issue. In this research study, we have implemented a system to systematically address the challenges in performing automatic music transcription.
item: Conference-Full-text
Framework to migrate angularjs based legacy web application to react component architecture
(IEEE, 2021-07) Kaushalya, T; Perera, I; Adhikariwatte, W; Rathnayake, M; Hemachandra, K
The next-generation component-based front-end frameworks such as React have outperformed AngularJS from both the end-user and developer expectations. The concepts, patterns, and architecture of the old school AngularJS application and the component-based React applications are different, hence the migration is not straightforward. This research defines a clear, cost-effective and efficient migration path from AngularJS applications to React applications. By comparing different approaches, the research provides a framework with a step-by-step guideline for migrating the presentation layer of any AngularJS based legacy application to a component-based React application. A migration assistant tool named Ng-React Copilot was implemented to guide the developer and automate the migration. The tool was developed by converting the critical refactorings proposed in the framework into a set of detection algorithms and providing the ability to scan against the given codebase. The tool works as a command-line tool as well as an integrated tool with popular IDEs. The framework and the tool were evaluated by applying them to selected small, medium, and enterprise-level AngularJS legacy applications and the results indicate the validity of the research findings.
item: Conference-Full-text
Miyaesi: java based implementation for automatic music transcription
(Computer Science & Engineering Society c/o Department of Computer Science and Engineering, University of Moratuwa., 2011-11) Abeykoon, H; Kaushalya, T; Akram, N; Dissanayaka, A; Weerawarana, S; De Silva, C; Weerawardhana, S; Madusanka, A; Dilrukshi, T; Aravinda, H
Music Notes play a major role in the music world. They are extremely important for musicians and composers. Sometimes people like to know music notes of an already composed music which raises the need of music transcription. In the past decade up to now researches and engineers have come up with vivid techniques to do music transcription automatically using probability and signal processing. With the advent of computer science which facilitated encoding and recording music in digital format it became an important topic. Nevertheless, implementation specific details are still rare to find addressing Automatic Music Transcription (AMT). In this paper we discuss concepts behind automatic music transcription and how they are applied in the system Miyaesi, an automatic music transcription system implemented using Java programming language. Further, we discuss how time domain signal analysis and spectrum analysis leads to automatic instrument identification.