Chun-Yu Lin

World models no longer require fragile offline reconstruction pipelines—baking native physics, depth, and camera pose directly into a unified multimodal generative process unlocks self-calibrating, closed-loop 3D spatial simulation at scale.

Achieve unprecedented spatial control in single-image generation by injecting 3D positional encodings into a diffusion model, enabling precise manipulation of object placement and scale.

Forget painstakingly curating 3D-consistent datasets: this RL approach uses a 3D foundation model to *verify* consistency, turning 3D scene editing into a tractable reinforcement learning problem.