Ouroboros: Single-Step Diffusion Models for Cycle-Consistent Forward and Inverse Rendering

While multi-step diffusion models have advanced both for- ward and inverse rendering, existing approaches often treat these problems independently, leading to cycle inconsis- tency and slow inference speed. In this work, we present Ouroboros, a framework composed of two single-step dif- fusion models that handle forward and inverse rendering with mutual reinforcement. Our approach extends intrin- sic decomposition to both indoor and outdoor scenes and introduces a cycle consistency mechanism that ensures co- herence between forward and inverse rendering outputs. Experimental results demonstrate state-of-the-art perfor- mance across diverse scenes while achieving substantially faster inference speed compared to other diffusion-based methods. We also demonstrate that Ouroboroscan transfer