Hong Kong University

From Denoising to Refining: A Corrective Framework for Vision-Language Diffusion Model

Discrete Diffusion VLA: Bringing Discrete Diffusion to Action Decoding in Vision-Language-Action Policies