As deep learning (DL) systems become integral to safety-critical domains, from autonomous driving to healthcare, their trustworthiness is more important than ever. Despite remarkable progress, deep neural networks can exhibit instability and remain vulnerable to adversarial perturbations. Addressing these issues requires new techniques that provide robustness guarantees and practical scalability. ...

Fibroblasts and the extracellular matrices they contribute support many different cell and tissue types in the body. They are also indispensable for tissue repair and for the regeneration of organs and limbs. Prof. Rinkevich is making significant contributions to the how fibroblast populations contribute to wound healing, cancerous growth, and tissue development. He is also making ground-breaking...

Generative AI (GenAI) has become an indispensable collaborator in various domains and reshaped how humans accomplish diverse tasks, ranging from data analysis to content creation. However, the current human- GenAI interaction loop faces three fundamental breakdowns: 1) the Instruction Gap, where users struggle to express complex and abstract intent with low-bandwidth prompts, 2) the Interpretation...

Given the scale and complexity of today's datasets, people face a critical challenge during visual data analysis: an overwhelming number of decisions. These decisions—about what, where, when, and how to visualize data—directly influence the quality of the analysis process and outcomes. For example, a user must decide what attributes to visualize, whether to apply filters, or which chart type best ...