The research team, including the University of Hong Kong and Zhiyuan, proposed a full-body VLA framework for humanoid robots.

On December 17, according to the news from Zhiyuan AGIBOT, a joint research team from the University of Hong Kong, Zhiyuan AGIBOT, Fudan University, and Shanghai Creative Academy proposed a WholeBodyVLA, a Vision-Language-Action framework for humanoid robots in the real world. The work, based on the release of the Zhiyuan Spirit X2 research, extends VLA to full-body control of bipedal humanoid robots, verifying its feasibility in whole-body loco-manipulation tasks.