(Ben) THON Pun Liang

I hold a Master of Science (Research) from Swinburne University of Technology Sarawak (2024), where I specialized in Explainable Artificial Intelligence (XAI) and Deep Learning for Medical Image Analysis. Currently, I work as a Data Scientist / AI Engineer in the Semiconductor Industry, focusing on anomaly detection, virtual metrology, and workflow automation.

I earned my Bachelor of Computer Science (Data Science) from the same university in 2021.

plthon [at] outlook [dot] com  /  LinkedIn  /  GitHub  /  Google Scholar /  Résumé

profile photo
Publications
Investigation of ConViT on COVID-19 Lung Image Classification and the Effects of Image Resolution and Number of Attention Heads
Pun Liang Thon, Joel C. M. Than, Norliza M. Noor, Jun Han, Patrick Then
IJIE, 2023
fullpaper

Convolutional Vision Transformer (ConViT) achieves high accuracy and improved performance in diagnosing COVID-19 cases from lung CT scan images, offering a valuable complement to RT-PCR tests.

Explainable COVID-19 Three Classes Severity Classification Using Chest X-Ray Images
Pun Liang Thon, Joel C. M. Than, Rosminah M. Kassim, Ashari Yunus, Norliza M. Noor, Patrick Then
IECBES, 2022
fullpaper

This study explores the application of vision transformers for COVID-19 severity classification using CXR images, achieving an accuracy of 0.862 and utilizing Grad-CAM for interpretability.

Preliminary Study on Patch Sizes in Vision Transformers (ViT) for COVID-19 and Diseased Lungs Classification
Joel C. M. Than, Pun Liang Thon, Omar Mohd Rijal, Rosminah M. Kassim, Ashari Yunus, Norliza M. Noor, Patrick Then
NBEC, 2021
fullpaper

This study investigates the potential of Vision Transformers (ViT) for classifying COVID-19 and normal lung images, achieving a promising peak accuracy of 95.36% across diverse datasets and patch sizes.

Misc

I served as a reviewer for BIBE 2022.



Last Updated: May 2025

Template