About Me
I'm Tharindu Wickremasinghe, a third-year PhD student at Purdue ECE, under the supervision of Prof. Stanley H. Chan at the Intelligent Imaging Lab. Prior to graduate school, I received my Bachelor's degree in Electronics and Telecommunications Engineering from the University of Moratuwa, Sri Lanka.
My research is in grounding powerful vision and language models in the physical laws and constraints of the real world, while also giving a human the ability to better interact and control the model. I focus on image/video enhancement using generative AI, working at the intersection of generative modeling, image/video restoration, and computational imaging.
I believe the power of AI is only effective if it is better controlled and interactive with a human user. Following this philosophy, I have contributed to research directions including FlowSteer, SeeU, and NewtonGen.
I am actively searching for internships for Summer 2027.
I'm always happy to collaborate and make new friends — feel free to reach out!
My hosts throughout the years:
News
| January 2026 | NewtonGen: has been accepted by ICLR 2026!. FlowSteer has been accepted by CVPR 2026(Findings), and SeeU has been accepted by CVPR 2026! |
| November 2025 | We released FlowSteer: which restores image degradations by steering flow models to preserve high pixel-level fidelity. |
| November 2025 | We released SeeU: a 4D understanding and video generation pipeline that reveals unseen time and space. |
| September 2025 | We released NewtonGen: a physics-consistent and controllable text-to-video framework. |
| April 2025 | Present my work with the IEEE-MEUYR program at ICASSP 2025- Hyderabad. |
| August 2024 | Started as graduate school as a graduate teaching assistant at Purdue University. TA for ECE301 Signals and Systems. |
| December 2023 | Graduated from the Faculty of Engineering, University of Moratuwa, with a First Class Honors in B.Sc (Engineering), Gold medals: for academics, and overall performance in leadership, and civic service. |
| November 2023 | Remote project with prof. Marco F. Duarte accepted for the "IEEE ME-UYR" program by the IEEE Signal Processing Society. |
| August 2023 | Invited to participate in the "Cornell-Maryland-MaxPlanck Research School" by at MPI Saarbruecken, Germany. |
| July 2023 | Joined the University of Moratuwa as a Visiting Instructor. |
| June 2023 | Presented my final year thesis project to the Department of Electronic and Telecommunications Engineering at the University of Moratuwa. |
| May 2023 | "MOSAIC": Masked Optimisation with Selective Attention for Image reconstruction" is published on arxiv. |
| November 2022 | Joined Centre for Advanced Imaging at Harvard University as a Remote Undergraduate Research Fellow , under supervision of Dr. Dushan. N. Wadduwage from Wadduwage Lab, focusing on deep learning based image reconstruction algorithms for microscopy. |
| January 2022 | co-authored paper: HPGNN: Hierarchical Point GNN for Outdoor Point Cloud processing accepted at ICPR 2022. |
| December 2021 | Joined the School of Computer Science, University of Sydney as a Research Intern. |
| July 2021 | Virtually attended ICASSP 2021 to defend our solution to the IEEE Signal Processing Cup, and Won the global championship |
| May 2021 | Appointed as Assistant Manager of the Knowledge Hub of the Electronic Club. |
| October 2018 | Started my undergraduate degree in Electronic and Telecommunications Engineering at University of Moratuwa, Sri Lanka. |
Research
Generative Images/Video
Computational Imaging
MOSAIC: Masked Optimization with Selective Attention for Image Reconstruction
P. Somarathne, T. Wickremasinghe, A. Niwarthana, A. Thieshanthan, C. Edussooriya, D. Wadduwage
arXiv preprint, 2023 arXiv
Embedded Vision & Autonomous Systems
HPGNN: Using Hierarchical Graph Neural Networks for Outdoor Point Cloud Processing
A. Thieshanthan, A. Niwarthana, P. Somarathne, T. Wickremasinghe, R. Rodrigo
ICPR 2022 proceedings
Towards Real-time Traffic Sign and Traffic Light Detection on Embedded Systems
O. Jayasinghe, S. Hemachandra, D. Anhettigama, S. Kariyawasam, T. Wickremasinghe, C. Ekanayaka, R. Rodrigo, P. Jayasekara
IEEE Intelligent Vehicles Symposium (IV) 2022 proceedings
Wireless Communications
Max-Min Fairness for IRS-Assisted Secure Two-Way Communications
Harindu Jayarathne, Tharindu Wickremasinghe, Kasun T. Hemachandra, Tharaka Samarasinghe, Saman Atapattu
IEEE Wireless Communications and Networking Conference (WCNC) 2025 proceedings
Previous research areas: Structured sparsity, Compressed Sensing and Inverse Problems in Imaging, Graph Neural Networks, Intelligent Reflective Surfaces, and Human Motion Prediction.
Through the IEEE SPS Young Researcher initiatives, I have collaborated with Prof. Marco F. Duarte on structured sparsity in images. I completed my Bachelors in Engineering with a First Class Honors and the gold medal for my cohort at the Department of Electronic and Telecommunication Engineering (ENTC), and the gold medal from the Faculty of Engineering, University of Moratuwa, Sri Lanka. My final year thesis project was on "Optical microscopy optimisation through compressive sensing and reconstruction" under the supervision of Dr. Dushan Wadduwage and Dr. Chamira Edussooriya. This research project was part of a collaboration with the Wadduwage Lab at the Centre for Advanced Imaging at Harvard University. I have worked on physics-based human motion prediction at the University of Sydney, Australia during my internship with Dr. Anusha Withana. Further, as self-initiated research projects, I have worked on point cloud segmentation using Graph Neural Networks with Dr. Ranga Rodrigo, and channel estimation and RIS configuration search with Dr. Prathapasinghe Dharmawansa.