Improving 3D Gaussian Splatting’s Rasterization Performance
After the completion of my software engineering research work, I found the opportunity with Professor Gilbert Bernstein to work on computer graphics research. In this project, I focused on finding algorithmic deficiencies in the rasterization process of 3D Gaussian Splatting, a novel technique for rendering 3D scenes. My goal was to improve the performance of this rendering method, which is crucial for real-time applications in computer graphics. Like my previous research, this project allowed me to apply my computer science knowledge in a practical context, except this time it was in the field of computer science I wanted to pursue in the future. My work here not only will deepen my understanding of computer graphics but will also demonstrate my ability to contribute to cutting-edge research in this area which is something I will need to apply to graduate school.
I made this research project the focus of my departmental honors senior thesis which I presented on in Spring 2025. I will continue to work on this project through the Summer of 2025 and work my way towards a publication in the Fall of 2025. The existence of this project and my work on it is a testament to my research journey at UW from my early aspirations for graphics research from the start to finally being able to work on it in a meaningful way and also be prepared to contribute quickly from all of my previous research experiences.