Hi, I am a Ph.D. Candidate in the Department of Computer Science at Stony Brook University (SBU), where I am fortunate to be advised by Prof. Dimitris Samaras. My research interests lie in the field of computer vision, machine learning and their applications in computational pathology. Previously, I earned my Master in Computer Science from Institute of Computing Technology, Chinese Academy of Sciences (ICT-CAS), where I focused on bioinformatics algorithms under the supervision of Prof. Prof. Shiwei Sun and Prof. Dongbo Bu. I got my Bachelor in Computer Science and Technology from Shandong University.

Research

My research interests lie in the field of multi-modal computer vision (CV), machine learning (ML) and their applications in multi-modal computational pathology. It is centered on three core pillars aimed at making deep learning models more practical and scalable for high-stakes applications:

• Efficient Network Training Schemes: I focus on developing strategies for training on large-scale images while navigating limited GPU memory constraints. This includes leveraging parameter-efficient tuning (e.g., prompt and lora), designing end-to-end training pipelines for Giga-pixel images, designing disentangled attention mechanisms to handle complex, multi-channel biological data.

• Efficient Network Architecture: I explore the design of linear-time models, specifically State Space Models (SSMs), which I optimize through custom CUDA kernels and hardware-aware operators for high efficiency. My work also incorporates disentangled attention and sparse attention strategies in image or multi-modal setting, such as attention sampling, to selectively process informative regions in gigapixel images.

• Self-supervised Learning: I am interested in advancing self-supervised learning for dense prediction tasks like segmentation and graph-based representations, enabling robust feature extraction from unlabeled data.

These research projects endow me with solid knowledge and rich experience in general deep learning and multi-modal computer vision models. I also has experience in high-performance GPU kernels. My research work are published in top-venue conferences and journals, including CVPR, TMLR, IPMI and MICCAI.

Collaboration & Mentorship

I believe in the power of collaborative research and have been fortunate to work closely with talented peers and faculty. I maintain a long-standing and productive collaboration with Saarthak Kapse and Xi Han, with whom I have co-authored several key works in multi-instance learning and efficient architectures. I also deeply value the interdisciplinary insights provided by our clinical collaborators, including pathologists Dr. Rajarsi Gupta and Dr. Vincent Quoc-Huy Trinh.

Beyond the guidance of my Ph.D. advisor, my academic trajectory has been significantly shaped by the mentorship and external support of several collaborating faculty members. I am deeply grateful for the intellectual contributions and guidance provided by Prof. Maria Vakalopoulou, Prof. Mahdi S. Hosseini, Prof. Prateek Prasanna and Prof. Hong Qin.

Personal

Outside of research, I am an active table tennis player and fitness enthusiast who loves exploring National Parks and hiking trails. I also enjoy diving into the creative world of ACG (Anime, Comic, and Games) as a fan of “2D culture.”