重點導讀
NVIDIA 發布開放世界動作模型 GR00T 2,可用於機械手臂、機器人手與人形機器人。
- 預訓練資料:包括 402,730 小時的人類頭戴攝影機影片、10,184 小時的手持夾爪示範,以及 23,500 小時的遙操作機器人資料。
AI 伴隨問答
針對這篇報導有疑問?可直接點選快捷問題或向 AI 追問背景脈絡與技術細節:
推薦理由
GR00T 2 的預訓練資料包含大量人類頭戴攝影機影片,並搭配較少的遙操作機器人資料,呈現其資料來源的組成。
Rohan Paul@rohanpaul_ai
·12 天前
Nvidia released GR00T 2, an open world-action model that runs arms, robot hands and humanoids.
Robots normally learn from recordings that pair each camera frame with exact motor commands, and those only come from a person steering a real robot.
Ordinary human video is far cheaper and more plentiful, but it has no motor commands, so earlier models could barely use it.
GR00T 2 eases that by pretraining on 402,730 hours of human head-camera video plus 10,184 hours of handheld-gripper demos, but only 23,500 teleoperated robot hours.
推文來源 · X (Twitter)在 X 參與討論
來源: Rohan Paul (x.com)
在 X 查看推文