Skip to content
RobotWorld
← Back

NaP-Control: Diffusion Prior for Fast Character Control

Loading video
Loading video

NaP-Control from ETH Zurich combines reinforcement learning with a task-agnostic diffusion motion prior, using latent noise optimization for precise and fast whole-body character control across goal reaching, velocity control and object interaction.

NaP-Control角色控制扩散先验强化学习全身控制物理仿真character controldiffusion priorreinforcement learningETH ZurichIsaacGymhumanoidTwitter
Category: humanoid
Author: @rsasaki0109
Date: 2026-08-23T00:00:00
Duration: 16.2s
Source: https://x.com/rsasaki0109/status/2091421974488924321
Reference: https://github.com/chiawenchen/NaP

Related Videos

Figure 03 Walking: The Helix 02 Three-Layer Control Architecture
1:05

Figure 03 Walking: The Helix 02 Three-Layer Control Architecture

@techniahqrobot · 43 views · 2026-08-27
人形机器人FigureHelix
CMU open-sources the HTD whole-body controller for humanoids — playable in your browser
1:17

CMU open-sources the HTD whole-body controller for humanoids — playable in your browser

@yaru_niu · 17 views · 2026-08-25
人形机器人全身控制Isaac Lab
PHYBOT: robot soccer as a full-contact locomotion test
0:51

PHYBOT: robot soccer as a full-contact locomotion test

@ErenChenAI · 13 views · 2026-08-15
PHYBOT机器人足球人形机器人
Robot's Rising-to-Stand Motion Demo
0:05

Robot's Rising-to-Stand Motion Demo

@XueJia24682 · 16 views · 2026-08-13
人形机器人humanoid起身站立
NVIDIA open-source humanoid controller training framework
0:57

NVIDIA open-source humanoid controller training framework

@neural_avb · 9 views · 2026-08-09
人形机器人NVIDIA强化学习
Physics-Based Character Controllers with RL
1:54

Physics-Based Character Controllers with RL

@neural_avb · 5 views · 2026-08-09
强化学习物理角色控制RL