Research program, State Key Laboratory of Nutrient Use and Management, CAU
Research program, Key Laboratory of Plant-Soil Interactions of MOE, CAU
In this research project, I delved into the paddy soil microbiome, specifically exploring how increasing salinity affects distinct microbial populations. Using metatranscriptomics with a double-RNA approach, I analyzed the metabolic responses of soil microorganisms to changing salinity levels, contributing valuable insights to sustainable agriculture practices.
Course project on Artificial intelligence and experiments (Grade: A)
My research focuses on using UNet and UNet++ architectures for bladder wall segmentation and tumor detection in MRI scans. UNet and UNet++ are renowned for their image segmentation capabilities.
UNet merges low-level and high-level features, enhancing segmentation. UNet++ further improves UNet with nested skip connections for detailed segmentation.
During UNet++ training, I faced gradient vanishing issues due to early stopping. UNet++ has more parameters than used in this project, and reducing them with pruning techniques remains a challenge.
Cornell Hackathon/CAU “Xingnong Cup”
During my participation in the Cornell University Digital Marathon, I led the development of an advanced crop model, earning a second-class award in the Xingnong Cup (CAU). This model integrates real-time monitoring and future prediction data, providing precise crop growth projections and yield forecasts.
My role involved providing crucial agricultural technical support and leading front-end development efforts. The project has secured software copyright, highlighting its potential to enhance agricultural practices and reduce production costs.
Undergraduate Research Project
This research integrates soil properties, uncovering the microbial aspects of carbon and nitrogen cycles. It enhances our knowledge of sustainable rice agriculture and informs decisions on carbon and nitrogen management.