PhD Candidate · University of Sydney

Tharinda
Piyadasa

AI & Machine Learning Researcher

I build systems that see, hear, and understand. My research sits at the intersection of medical imaging, speech science, and deep learning. Currently developing 3D vocal tract models from real-time MRI using diffeomorphic deformation and acoustic simulation.

View Research Download CV
Tharinda Piyadasa

Research Interests

Computer Vision Speech & Acoustic Analysis Medical Imaging Multimodal Learning Diffeomorphic Registration

Experience

Jun 2023 — Present

Doctoral Researcher & Casual Academic

University of Sydney · Sydney, Australia

  • 3D vocal tract modelling from real-time MRI using constrained LDDMM and diffeomorphic deformation.
  • Deep learning models for automated vocal tract analysis and acoustic feature extraction.
  • 6 published papers including 3 at Interspeech 2025; collaborating with an interdisciplinary team of 10+ researchers.
  • Teaching assistant for DATA2001, COMP9123, INFO1110, INFO5992, and INFO6007.

May 2022 — May 2023

AI Research Engineer

BooleanLabs · Sri Lanka

  • Developed a real-time suspiciousness detection system integrating YOLOv4-based detection.
  • Built a deep learning fabric recommender system using multi-model fusion and image feature extraction.

Oct 2020 — Apr 2021

Software Engineer Intern

WSO2 LLC · Sri Lanka

  • Automated the Ballerina Standard Library CI/CD build pipeline, reducing manual effort and improving release reliability.
  • Designed the Ballerina SerDes package using Google Protocol Buffers (Proto3) for efficient typed data transfer.

Education

Jun 2023 — Present

PhD, Electrical and Computer Engineering

University of Sydney · Australia

3D vocal tract modelling using real-time MRI, diffeomorphic deformation, deep learning segmentation, and acoustic validation.

2018 — 2022

B.Sc. (Hons) in Computer Science — GPA 3.99 / 4.0

University of Colombo · Sri Lanka

Prof. Mohan Munasinghe Award — Highest GPA in Computer Science.