As a research assistant in the Machine Learning, Perception, and Cognition Lab at UCSD, I work under the supervision of PhD student Zeyuan Chen and Professor Zhouwen Tu. My research focuses on deep learning techniques for image processing, where I have reproduced results from state-of-the-art papers like Re-Confusion, PixelSplat, and NeRF. These projects have provided me with practical experience in implementing novel techniques and debugging complex models.
Recreating State-of-the-Art Results
In my role, I’ve implemented deep learning imagery techniques based on the latest research in the field. I’ve successfully reproduced results from several papers and contributed to improving the methods for future applications. These projects have allowed me to develop a strong understanding of how to work with complex image data in PyTorch.
Debugging & Dependency Management
A key part of my responsibilities is debugging issues that arise with PyTorch libraries and fixing problems related to mismatched dependencies. I’ve developed strong skills in managing these technical challenges, ensuring the smooth running of experiments and models.
Novel Research Development
Currently, I am collaborating with the lab team on creating novel deep learning methods, which we aim to publish in top conferences.