this is
My portfolio
I have had hands-on experience in creating computer vision, NLP and AI related applications.

I'm a fresh graduate (officially graduating in December 2022) from Department of Electronic and Telecommunication Engineering,
University of Moratuwa,
Sri Lanka.
I am an individual who likes to solve challenging problems
utilising analytical skills and engineering principles.
Interested in Software Engineering, Computer Vision, NLP and Web/App Development.
I have had hands-on experience in creating computer vision, NLP and AI related applications.
An intelligent retrofittable monitoring system that is capable of tracking activity,
saving energy,predicting and detecting faults with its type in commercial cooolers.
Web platform for reputation management in social medias powered by Aspect based sentiment
analysis to track the customer satisfaction
and attitude towards the business.
Conducted a research on deepfake generation and detection, then integrated a deepfake detection model
to a face recognition platform at zone24x7.
A web application for classifying sentiments of customer reviews into positive,
negative and neutral with scores and highlighting.
This is an ALPR (Automatic License Plate Recognition) framework that is improved
for cpu inferencing with ORT (OnnxRunTime) and capable of detecting,recognizing and blurring vehicle
license plates.
A Desktop app for detecting highway lane violations realtime powered by OpenVino. Illegal parking,driving at wrong direction,
and improper overtaking can be detected.
Conducted research on age estimation, trained a model and updated age estimation model
in Cognative Visual Analytic Platform (CVAP) at Zone24x7.
A human detection, tracking and counting system built using deep learning based computer vision
powered by pedestron and SORT (Simple, Online, and Realtime Tracking) algorithm
A facial recognition system with a GUI which uses a live feed (web cam)
and recognizes the person real-time. When new users are added, system will
automatically train and update the model.
Monitoring queues to check whether people are following the one meter
distance social distancing rule and if violated corresponding people will be
addressed by their line numbers with an audio.
Hand recognition system is a biometric system which
uses the hand geometry and the palm print to recognize a person. Then unique
features from the palm print are extracted to create an embedding for
the palm. With that system tries to recognize the person.
A desktop app for monitoring exercises of rheumatoid arthritis
patients using pose estimation. Given a exercising video as the input, the app
will zoom out on the important joints so the doctors can monitor the procedure well.
Designing and Implementing a multi-core processor to
multiply two matrices of any dimension.
Designing and Building a Lux Meter that can be used to measure the
luminocity of the environment.