Skip to content

Research

Page 28 of 166

Masked Visual Actions for Unified World Modeling

research note

Masked Visual Actions for Unified World Modeling

·8 min read·Hadi Alzayer, Wenlong Huang, Haonan Chen et al.

This paper addresses the challenge of integrating low-level robot actions into pretrained video models to enable unified forward and inverse robotic world.

researchrobotic-world-modelingmasked-visual-actionsvideo-diffusionpixel-aligned-conditioning

Read note → Source paper ↗

Articles are CC BY 4.0 — feel free to quote with attribution