Zi Ye
| yezi27@mail2.sysu.edu.cn |
|
| +86 13318338110 |
|
Education
- 2021-2025 (expected)
-
BEng, Artificial Intelligence; Sun Yat-sen
University
- Thesis title: Research on GPU-Optimized Fast Retrieval Model for
Medical Image Vector Data
- Advisor: [Assoc. Prof. Libin Zheng]
Research Experience
Evolutionary Multi-task Optimization for Multi-task
Reinforcement Learning
- Participated in a five-member team to apply evolutionary algorithms
to multi-task deep reinforcement learning.
- Conducted a literature review on evolutionary-guided RL methods,
identifying key limitations in cross-task optimization.
- Implemented MFEA and A-MFEA-RL in Python to evolve multiple deep
Q-learning models simultaneously.
- Advisor: Assoc.
Prof. Zefeng Chen
Research on GPU-Optimized Fast Retrieval Model for Medical
Image Vector Data
- Conducted a comprehensive review of large-scale vector retrieval
methods.
- Introduced an ROI-scoring mechanism that combines ResNet-based
disease classification with U-Net segmentation.
- Developed a class-weight–aware scheduling strategy and
cache-eviction policy that prevents rare-disease clusters from being
evicted prematurely.
- Advisor: Assoc.
Prof. Libin Zheng
Research Interest
- Vision-Language Models for Remote Sensing
- Multimodal Meteorological Forecasting
- Multi-Agent Collaboration with VLMs