Yezhou Yang

Yezhou (YZ) Yang is a professor in the School of Computing and Artificial Intelligence at Arizona State University whose research centers on vision-and-language intelligence, generative AI, and visual reasoning. His work explores how machines learn and compose visual concepts for image and video understanding, with recent emphasis on controllable and secure generative models, diffusion and flow-based generation, and concept-level learning. He has contributed to benchmarks and methods for instruction-based image editing, action depiction in text-to-image systems, video question answering, and event-based perception, as well as studies on hallucination and interpretability in vision–language models from cognitive perspectives. His research spans both foundational modeling—such as compositional image generation, concept erasure, and alignment—and applied systems for localization, traffic understanding, and cooperative perception. His publications appear in leading venues including NeurIPS, ICCV, EMNLP, IJCAI, CVPR workshops, and WACV, reflecting a strong interdisciplinary focus on robust, interpretable, and trustworthy visual intelligence.