To address this, we propose the VLA model H-RDT (Human to Robotics Diffusion Transformer), which is pretrained on a ...