Find Tharindu Fernando Warnakulasuriya on
Tharindu Fernando Warnakulasuriya's full profile, contact information and publications are available on:
Doctor of Philosophy (Queensland University of Technology)
Dr. Tharindu (Fernando) Warnakulasuriya is a Research Fellow in the Signal Processing, Artificial Intelligence, and Vision Technologies (SAIVT) research program in the school of Electrical Engineering and Robotics at Queensland University of Technology (QUT). Tharindu received his BSc (special degree in computer science) from the University of Peradeniya, Sri Lanka, and his PhD from QUT, Australia. His PhD thesis was entitled “Context Modelling for Single and Multi-Agent Trajectory Prediction” where he developed systems to predict future behaviour in single and multi-agent settings. The underpinning theoretical innovation of his thesis was the development of a novel Neural Memory Network (NMN) architecture to capture both short-term and long-term context of a given domain. He was awarded a QUT Outstanding Doctoral Thesis Award in recognition of ground-breaking innovative research. Since the completion of his PhD, Tharindu is conducting interdisciplinary research activities and collaborating with researchers in healthcare, neuroscience, psychology, and computer vision to solve challenging problems in several domains, utilising the advances that he contributed to the area of Machine Learning during his PhD and beyond. Research highlights:
- Neural memory plasticity for medical anomaly detection.
- Automated system for Terminal Area Air Traffic Prediction.
- Algorithm to Predict Tennis Players' Next Shots.
- Computer-Aided System for Abnormal Event Detection from Surveillance Feeds.
- Autonomous Steering and Driver Behaviour Prediction for Autonomous Cars.
- Monitoring Intuitive Expertise of Airport Security Screeners.
Projects
Additional information
Research Areas
- Machine Learning: Deep Learning, Recurrent Neural Networks, Neural Memory Networks, Interpretable Machine Learning, Generative Adversarial Networks, Generative Adversarial Imitation Learning, Deep Inverse Reinforcement Learning, Multimodal Deep Learning.
- Biomedical Signal Processing: Abnormality Detection, Medical Imaging, Biosignal Classification.
- Computer Vision: Human Behaviour Analysis and Prediction, Person Tracking, Saliency Prediction, Image Forensics.
- Sports Analytics: Player Behaviour Analysis and Prediction.
Research Applications:
- Machine Learning: Neural Memory Plasticity, Deep Context Modelling, Structured Memory Networks, Attention Driven Fusion, Recurrent Attention Networks.
- Biomedical Signal Processing: Heart State Segmentation, Abnormal Heart Sound Detection, Automated Schizophrenia Risk Detection.
- Computer Vision: Pedestrian Trajectory Prediction, Aircraft Trajectory Prediction, Monitoring Intuitive Expertise of Airport Security Screeners, Pedestrian Group Detection, Autonomous Steering, Discovering Hidden Temporal Patterns, Human Action Recognition.
- Sports Analytics: Next Shot Prediction in Tennis, Player Strategy Analysis, Soccer Event Analysis.
Professional Service
- Reviewer: IEEE Transactions on Information Forensics and Security; IEEE Transactions on Neural Networks and Learning Systems; Journal of Biomedical and Health Informatics; IEEE Transactions on Circuits and Systems for Video Technology.
- Type
- Academic Honours, Prestigious Awards or Prizes
- Reference year
- 2019
- Details
- 2019 QUT University Award for Outstanding Doctoral Thesis
- Type
- Academic Honours, Prestigious Awards or Prizes
- Reference year
- 2015
- Details
- University Award for Academic Excellence, University of Peradeniya, Sri Lanka
- Fernando, T., Gammulle, H., Denman, S., Sridharan, S. & Fookes, C. (2022). Deep Learning for Medical Anomaly Detection: A Survey. ACM Computing Surveys, 54(7). https://eprints.qut.edu.au/214059
- Thilakasiri Dissanayakage, P., Warnakulasuriya, T., Denman, S., Sridharan, S. & Fookes, C. (2022). Geometric Deep Learning for Subject Independent Epileptic Seizure Prediction using Scalp EEG Signals. IEEE Journal of Biomedical and Health Informatics, 26(2), 527–538. https://eprints.qut.edu.au/212250
- Fernando, T., Sridharan, S., Denman, S. & Fookes, C. (2022). Split ‘n’ merge net: A dynamic masking network for multi-task attention. Pattern Recognition, 126. https://eprints.qut.edu.au/232375
- Fernando, T., Fookes, C., Denman, S. & Sridharan, S. (2021). Detection of Fake and Fraudulent Faces via Neural Memory Networks. IEEE Transactions on Information Forensics and Security, 16, 1973–1988. https://eprints.qut.edu.au/210209
- Gammulle, P., Warnakulasuriya, T., Denman, S., Sridharan, S. & Fookes, C. (2019). Coupled generative adversarial network for continuous fine-grained action segmentation. Proceedings of the 2019 IEEE Winter Conference on Applications of Computer Vision (WACV), 200–209. https://eprints.qut.edu.au/126905
- Warnakulasuriya, T., Denman, S., Sridharan, S. & Fookes, C. (2019). GD-GAN: Generative adversarial networks for trajectory prediction and group detection in crowds. Computer Vision - ACCV 2018: 14th Asian Conference on Computer Vision, Revised Selected Papers, Part I (Lecture Notes in Computer Science, Volume 11361), 314–330. https://eprints.qut.edu.au/126868
- Warnakulasuriya, T., Denman, S., Sridharan, S. & Fookes, C. (2019). Pedestrian trajectory prediction with structured memory hierarchies. Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2018, Proceedings, 241–256. https://eprints.qut.edu.au/126867
- Warnakulasuriya, T., Denman, S., Sridharan, S. & Fookes, C. (2018). Soft + Hardwired attention: An LSTM framework for human trajectory prediction and abnormal event detection. Neural Networks, 108, 466–478. https://eprints.qut.edu.au/126862
- Denman, S., Sridharan, S., Fookes, C. & Warnakulasuriya, T. (2018). Learning temporal strategic relationships using generative adversarial imitation learning. Proceedings of the 17th International Conference on Autonomous Agents and MultiAgent Systems, 113–121. https://eprints.qut.edu.au/118094
- Warnakulasuriya, T., Denman, S., Sridharan, S. & Fookes, C. (2018). Task specific visual saliency prediction with memory augmented conditional generative adversarial networks. Proceedings of the 2018 IEEE Winter Conference on Applications of Computer Vision (WACV 2018), 1539–1548. https://eprints.qut.edu.au/118091
- Deep Learning on Multi-Modal Data
PhD, Associate Supervisor
Other supervisors: Professor Sridha Sridharan, Professor Clinton Fookes, Dr Simon Denman - Self-Supervised Learning for 3D Multimodal Perception
PhD, Associate Supervisor
Other supervisors: Professor Clinton Fookes, Adjunct Professor Peyman Moghadam, Professor Sridha Sridharan - Deep Spatial-Spectral Representation Learning for Hy-perspectral Data
PhD, Associate Supervisor
Other supervisors: Professor Sridha Sridharan, Adjunct Professor Peyman Moghadam, Professor Clinton Fookes - Generalized Interpretable Deep Learning Models for Biosignal Classification
PhD, Associate Supervisor
Other supervisors: Professor Clinton Fookes, Professor Sridha Sridharan, Dr Simon Denman