Zhicheng Xu
MS-GIST Student & Researcher
Georgia Institute of Technology
Atlanta, GA, USA
eet You!
Please call me Bruce or Katzemann
if you feel difficult to pronounce my Chinese name.
About
I am a graduate student in the MS-GIST program at Georgia Tech, with interests in the application of artificial intelligence and high-performance computing to spatial data science. My background, rooted in my work at Key Laboratory of GIS, Ministry of Education, East China Normal University, includes extensive research in remote sensing, computer vision, and spatial analysis, leading to a co-authored SCIE publication. I am also a passionate open-source contributor dedicated to advancing the field and sharing knowledge.
Actively seeking PhD/RA opportunities in GIS related area, as well as Machine Learning Engineer & GIS Engineer positions.
Education
Master of Science in Geographic Information Science and Technology (MS-GIST)
Aug 2025 - PresentGeorgia Institute of Technology, School of City and Regional Planning
Bachelor of Science in Geographic Sciences | Minor in AI + Math
Sept 2021 - Jun 2025East China Normal University (ECNU), School of Geographic Sciences
Publications
"Optimizing the Spatial Accessibility of Outdoor Sports Facilities: A Greedy Heuristic Algorithm Based on Remote Sensing Images"
International Journal of Digital Earth, 2024, 17(1)
DOI: 10.1080/17538947.2024.2390427
Most Recent Research Experience
Bachelor's Thesis: A Research on the Construction of Lightweight Open Vocabulary Detector for Remote Sensing Images
October 2024 – May 2025Supervised by Prof. Jianping Wu, Key Laboratory of Geographic Information Science, Ministry of Education, ECNU
- Introduced a light-weight solution for training open vocabulary detection models for remote sensing images using the xView dataset and a single RTX 4090 GPU
- Designed and implemented pretraining on xView, dataset transfer (zero-shot and fine-tuning) on DIOR, and task transfer fine-tuning on Potsdam dataset
- Explored and discussed the mechanism of the open vocabulary detection model with basic machine learning theory
- Technical report available on GitHub
Deep Learning-Based Recognition and Accessibility Analysis of Outdoor Stadiums in the Yangtze River Delta
May 2024 – May 2025Supervised by Dr. Shenjun Yao, Key Laboratory of Geographic Information Science, Ministry of Education, ECNU
- Introduced a slicing-aided pipeline for efficiently selecting data to label with CLIP zero-shot interference
- Trained an appropriate model for sports facilities in ultra-high resolution aerial images based on YOLO11-obb
- Constructed the complete workflow of training new object detection model, using SAHI for fine-grained detection, and digitalize the box to AOI data
- Modified the Gaussian 2SFCA method to a 3-step approach including an extra supply level step (using z-score norm)
- Selected as a "Municipal-level undergraduate innovation project" by Shanghai City
- Code available on GitHub and Kaggle
Selected Open Source Projects
My Kernel Templates
June 2025 - OngoingTechnologies: CUDA, Triton, GPU Kernels
A continuous code repository to store the kernels I implemented in Triton or CUDA. The Flash-Attention sub-project provides a pure Triton kernel for the forward and backward passes of Flash Attention-2. The related tech blog has received over 8k reads, 341 likes, and 384 collects on Rednote.
Repository: GitHub
How to Teach Your Cat DDPM
Oct 2024 - Dec 2024Technologies: DDPM, Machine Learning Theory, LaTeX
An interest-led project that reproduced the proof of the DDPM model from "Understanding Diffusion Models: A Unified Perspective" (arXiv:2208.11970, 2022) and practiced mathematical writing with LaTeX.
Repository: GitHub
Professional Experience
Teaching Assistant - Software Engineering and GIS Development
Feb 2024 - Jul 2024East China Normal University, Shanghai, China
- Tutored in the 36-hour theory class, organized regular Q&A sessions, implemented extended learning about the GDAL library and debugging
- Instructed in the 18-hour lab class, assisting 75 students in setting up environments, completing assignments, and final projects
Technical Skills
Programming Languages
Python, C++, MATLAB, C#, LaTeX, HTML/CSS
Frameworks & Tools
PyTorch, Hugging Face, Ultralytics, Triton, CUDA, FastAPI, Streamlit, Qt6
Software & Specializations
ArcGIS, QGIS, Computer Vision, Large Language Models, Deep Learning, Remote Sensing
Honors & Awards
2024 ICM (Interdisciplinary Contest in Modeling)
Honorable Mention (Problem D)2023 CUMCM
Third Prize, Shanghai Region2023 Geographic Information Technology Competition
Individual Third PrizeOutstanding Undergraduate Scholarship
ECNU, 2022-2024 (Three times)Outstanding Student Leader
ECNU, 2022-2023 & 2023-2024Leadership & Activities
President, Department Leader, Officer @Geography E+ Club (2021-2024)
ECNU School of Geographic Sciences
Student Mentor
1+8 Mentorship Program, ECNU
Volunteer
ECNU Mengyou Small Animal Protection Society
Personal Interests
Guitar & Music Writing, Tennis & Soccer, Algorithm Competitions, Digital Illustration, Simulate Racing
Languages
Mandarin (Native), English (IELTS: 7.5)