Wisdom O. Ikezogwo

My research asks what it takes to build intelligence trustworthy enough to act in high stakes visual domains such as medical imaging. I study multimodal representation and generative modeling: how models form rich, robust, and grounded connections across modalities and scales, learn from scarce or noisy supervision, reason under uncertainty, and improve through interaction. I care about the full arc, from curating the right data, to grounding what a model perceives, to reasoning over it, to closing the loop with feedback and action. Right now, my research interests are in visually grounded reasoning and in improving world/action models.