Skip to content
Agientry
ZH
← Back to Videos

NVIDIA Kimodo turns text and constraints into robot motion

Loading video
Loading video

NVIDIA Kimodo generates controllable 3D human and G1 robot motion from text, keyframes, end-effector constraints, and paths using 700 hours of optical mocap.

NVIDIAMotion GenerationDiffusion modelsHumanoidScenario
Category: humanoid
Author: @cutetoxicguy
Date: 2026-11-08T00:00:00
Duration: 5.0s
Source: https://x.com/cutetoxicguy/status/2119558081352729010
Reference: https://research.nvidia.com/labs/sil/projects/kimodo/

Related Videos

kimodo.cpp: NVIDIA Kimodo text-to-motion runs on CPU/Vulkan
0:38

kimodo.cpp: NVIDIA Kimodo text-to-motion runs on CPU/Vulkan

@Stefan_3D_AI · 828 views · 2026-08-23
kimodo.cppKimodo
RoboSkills Generates Humanoid Motion From Video
1:19

RoboSkills Generates Humanoid Motion From Video

@gengchen · 150 views · 2026-11-02
Agent WorkflowMotion Generation
MotionBricks: Real-Time Robot Motion in the Browser
0:40

MotionBricks: Real-Time Robot Motion in the Browser

@victormustar · 154 views · 2026-10-28
MotionBricksMotion Generation
GR00T 2 Learns a New Task From Five Demos
0:19

GR00T 2 Learns a New Task From Five Demos

@0x_meden · 53 views · 2026-11-10
HumanoidRobot Learning
Humanoid Agility and Whole-Body Control Reel
1:42

Humanoid Agility and Whole-Body Control Reel

@AlecSaysThings · 92 views · 2026-11-07
HumanoidMotion Control
Metatron: Goal-Driven Humanoid Motion Control
0:15

Metatron: Goal-Driven Humanoid Motion Control

@AlecSaysThings · 156 views · 2026-10-28
MetatronHumanoid

As an Amazon Associate, we earn from qualifying purchases.