Jin Sun

In my research, I care deeply about solving fundamental questions in computer vision and general AI, and about building on these results to create models and tools that push the boundary of science and improve people’s quality of life.

1. Computer Vision

   Core problems that shape how intelligent systems see, interpret, and generate the visual world, and how to build models that can do these things well.

   
   

   Generation

   Editing

   Object Recognition

   3D Geometry

   Lighting

   Context Modeling

   Human Analysis

   Visual Reasoning

2. Fundamental AI

   How deep neural networks behave, learn, generalize, and build useful multimodal representations.

   
   

   Neural Networks

   Multimodal

   LLM

   VLM

   LMM

   Generalization

   Interpretability

   Representation Learning

   Learning with limited data and labels

4. AI for Science and Society

   How to create meaningful impact through AI beyond traditional computer vision and machine learning applications.

   
   

   Agriculture

   BioScience

   Geography

   Health

   Public Safety

   Transportation

   Education