In my research, I focus on analyzing high-dimensional imaging data through deep learning techniques. My projects include: (1) Developing a universal, large-scale kidney nuclei segmentation model for kidney pathology microscopy. (2) Implementing active learning strategies to refine foundation models in kidney pathology. Particularly, improving the alignment between model prediction and human preferences. (3) Pioneering the training and deployment of the first DINOV2 foundation model for WVII and WVIII remote sensing satellite imagery. These initiatives are designed to provide the research community with robust AI foundation models in both kidney pathology and satellite imagery. These aspects align closely with VALIANT, an initiative dedicated to advancing AI research and its applications through collaborative efforts with domain experts.