Tech Meridian ← LIVE FEED
PROMY MERIDIAN RU

RELEASE · MODELS · #3367

Nvidia unveils GR00T 2 world-action model to transfer human video movements to robots

Nvidia's research report on GR00T 2 describes a 'world-action' foundation model that predicts both next robot actions and short future video frames, enabling transfer of human demonstrations from head-mounted videos to different robot embodiments. Trained on a cluster using over a thousand GB300 chips and about 400,000 hours of footage, the system (and a Turbo variant for bipedal robots running on Jetson Thor) achieved mixed task success rates (10–80%), rapid few-minute fine-tuning for some tasks, and higher assembly reliability when paired with a higher-level planner such as OpenAI's GPT-6 Astra, but still faces limitations like lack of tactile input, missed fine errors, and latency from the planner.

KEY POINTS

  1. Nvidia's research report on GR00T 2 describes a 'world-action' foundation model that predicts both next robot actions and short future video frames, enabling transfer of human demonstrations from head-mounted videos to different robot embodiments.
  2. Trained on a cluster using over a thousand GB300 chips and about 400,000 hours of footage, the system (and a Turbo variant for bipedal robots running on Jetson Thor) achieved mixed task success rates (10–80%), rapid few-minute fine-tuning for some tasks, and higher assembly reliability when paired with a higher-level planner such as OpenAI's GPT-6 Astra, but still faces limitations like lack of tactile input, missed fine errors, and latency from the planner.
  3. If robust, GR00T 2's video-prediction approach could cut costly teleoperation by letting robots learn tasks from human videos and transfer skills across different robot bodies, accelerating robot deployment.

WHY IT MATTERS

If robust, GR00T 2's video-prediction approach could cut costly teleoperation by letting robots learn tasks from human videos and transfer skills across different robot bodies, accelerating robot deployment.

SOURCES & TIMELINE

1