NVIDIA's GR00T 2 transfers human skills to robots with just five task demonstrations
NVIDIA’s GR00T 2 transfers human skills to robots with just five task demonstrations
NVIDIA introduced GR00T 2, a new generation of robot foundation models capable of transferring skills demonstrated by humans to machines that have never practiced those tasks.
Source: Robotics — Robot learning news discovery · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
Reported numbers
task demonstrations
5
View original evidence
NVIDIA’s GR00T 2 transfers human skills to robots with just five task demonstrations
Open source S1robot data per task
<3 minutes
View original evidence
95 percent following a 30-minute fine-tuning session on a single NVIDIA RTX 5090 graphics card. The demonstrations required less than three minutes of robot data per task. Robots learn from human demonstrations NVIDIA calls the approach cross-adaptation, or X-Adapt. It allows a robot to transfer skills from demonstrations performed using a different body or device. In one experiment, researchers showed a robot how t
Open source S4human video training hours
402,730 hours
View original evidence
he assembly in 60 percent of trials, despite never receiving a task-specific demonstration. Researchers also used a wearable device that mirrors the movements of a person’s fingers to teach a robot hand two-handed tasks. These included preparing banana yogurt and scooping granola, as well as striking a match to light a lamp. The model’s training combines 402,730 hours of first-person human video, 10,184 hours of hand
Open source S6handheld-gripper demonstration hours
10,184 hours
View original evidence
he assembly in 60 percent of trials, despite never receiving a task-specific demonstration. Researchers also used a wearable device that mirrors the movements of a person’s fingers to teach a robot hand two-handed tasks. These included preparing banana yogurt and scooping granola, as well as striking a match to light a lamp. The model’s training combines 402,730 hours of first-person human video, 10,184 hours of hand
Open source S6Human-1M dataset available hours
100,000 hours
View original evidence
held-gripper demonstrations and 23,500 hours of robot teleoperation data. NVIDIA says 100,000 hours of its Human-1M dataset are openly available. Reasoning models direct robot actions GR00T 2 can also work with reasoning models that break complex tasks into steps, monitor progress, and direct the robot through an application programming interface. In tests involving three toy assemblies, this approach completed at le
Open source S7
- paper: not_reported
- dataset: Reported pending Open source S7
- code: not_reported
- weights: not_reported
- dataset: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
NVIDIA’s GR00T 2 transfers human skills to robots with just five task demonstrations
Open source S1
95 percent following a 30-minute fine-tuning session on a single NVIDIA RTX 5090 graphics card. The demonstrations required less than three minutes of robot data per task. Robots learn from human demonstrations NVIDIA calls the approach cross-adaptation, or X-Adapt. It allows a robot to transfer skills from demonstrations performed using a different body or device. In one experiment, researchers showed a robot how t
Open source S4
he assembly in 60 percent of trials, despite never receiving a task-specific demonstration. Researchers also used a wearable device that mirrors the movements of a person’s fingers to teach a robot hand two-handed tasks. These included preparing banana yogurt and scooping granola, as well as striking a match to light a lamp. The model’s training combines 402,730 hours of first-person human video, 10,184 hours of hand
Open source S6
held-gripper demonstrations and 23,500 hours of robot teleoperation data. NVIDIA says 100,000 hours of its Human-1M dataset are openly available. Reasoning models direct robot actions GR00T 2 can also work with reasoning models that break complex tasks into steps, monitor progress, and direct the robot through an application programming interface. In tests involving three toy assemblies, this approach completed at le
Open source S7
umanoid . It runs directly on the Unitree G1 using an onboard Jetson Thor processor, with each prediction taking 124 milliseconds. Through a whole-body controller, the humanoid completed popcorn-serving tasks in all 48 trials and cleared a table of different objects in 93 percent of trials. NVIDIA introduced GR00T 2 at the Conference on Robot Learning in Austin, Texas, on November 9. The company says the model is int
Open source S9
Source：Robotics — Robot learning news discovery · msn.com