We develop human-centered tools that combine AI algorithms with interactive visualization to build digital models of visual aspects of the world and to explore large-scale datasets more intuitively. Moving away from "black box" systems, we develop AI tools that can integrate context and interact with users more naturally. Our research focuses on novel visual representations and methodology for capturing and representing data to support tasks like data cleaning, modeling, and hypothesis testing. By integrating these computational methods, we build systems that make AI more transparent and accessible for researchers and users.
Visual and Interactive AI
The Visual and Interactive AI (VIAI) research unit works at the intersection of computer graphics, AI-driven visualization, computer vision, and machine learning to address complex scientific challenges.
