I lead research on robot perception and control, generative 3D scene understanding, and human-robot interfaces. We combine reinforcement learning, vision-based localization, and multimodal language-vision models to enable reliable, user-guided autonomy for manipulation, navigation, and interactive imaging.