Stem-OB: Generalizable Visual Imitation Learning with Stem-Like Convergent Observation through Diffusion Inversion

The tree of Stem-OB inversion is composed of different objects progressively inverted through a diffusion inversion process. Moving downward alone the tree's branches, objects of different textures, appearances, and categories gradually get closer, eventually converging into the same root of Gaussian noise, where they are completely indistinguishable. Right: Real-world tasks success rate, where Stem-OB showcases a significant improvement.