‘One Model, One Data Interface, Any Body’: Reward AI’s OM-1 Learns Manipulation Straight From Humans
Covered by 1 source · 1 article
Reward AI unveils OM-1, a robot foundation model trained directly on human motion — no teleoperation or robot data needed — zero-shot generalizing across arms and humanoids. The post ‘One Model, One Data Interface, Any Body’: Reward AI’s OM…
Covered by
