14000776_dilumi

Dilumi Liyanage

A hardworking and versatile computer science undergraduate with a commitment to do the work assigned to me. I am good at doing team work and have excellent communication skills.


Email: liyanagedilumi@gmail.com
Phone: 0715563635
Course: CS

Interests: Natural Language Processing, Image Processing, Machine Learning, Data Analytics, Digital Forensics, R&D

Technical Skills: JavaEE (with EJB and JPA), JavaSE, Laravel, C, PHP, Python, Oracle SQL, MySQL R, MATLAB

Project Experience

JustPay Admin Module: A system that facilitates merchant & sub merchant maintenance, users &
transactions monitoring and report generation for inter-bank low value merchant payments
CEFT transaction facility between accounts and cards in ESB Services Parameter Maintenance App: This system has the functionality of adding new beneficiary banks to the CEFT system and enabling/disabling of CEFT mode for CEFT participating banks
Online password reset option for Sampath Vishwa Retail: Online banking solution provide
Student Management System: for BIT external degree program by University of Colombo School of Computing
Web based system for a salon: This system facilitates online ordering of cosmetic products and making appointments online

Work Experience

Software Engineering Intern | Sampath Bank PLC (August 2017 – February 2018)

Achievements | Awards

Microsoft Imagine Cup Sri Lanka – 2015 | Games Category – 2nd Runner-Up
Participation: IEEExtreme 9.0 | IEEExtreme 10.0
Participation: Hackerholicks hackathon 2016
Certificate in Business Accounting – CIMA
Diploma in English Aquinas University College

Final Year Project

Improving Sinhala OCR using Deep Learning

Sinhala Optical Character recognition performance is not good on unrestricted text such as identifying different fonts, similar characters and different font sizes. Languages such as English and Latin-based languages have achieved state-of-the art character recognition accuracies using Deep Learning. However, such approach has not been witnessed for printed Sinhala complex character recognition. Our attempt is to improve the performance of Sinhala printed character recognition on unrestricted text using Deep Learning.