At Shanghai University of Engineering Science, I assisted with the Mathematical Analysis course, supporting undergraduate students with topics including limits, differential calculus, and series. I also evaluated assignments and tests and supported faculty during final examinations.
For my Brain Tumor Classification from MRI Images project, I developed a lightweight CNN for four-class classification that achieved 94% accuracy and 0.99 recall. I reduced the model to 1.7 million parameters using Global Average Pooling.
I built a semantic image search system using OpenAI’s CLIP for natural-language queries over image archives. I integrated Qdrant Vector DB for similarity search and implemented zero-shot classification for automated image tagging.
On my Real-time Monitoring System project, I designed a concurrent architecture with a Spring Boot backend and a multithreaded socket server. I also implemented synchronization primitives and optimized JPA/Hibernate performance.
