VRSO introduces a visual-centric approach for static object annotation, providing high-quality annotations efficiently and accurately using only camera images as inputs.
