DreamControl: Human-Inspired Whole-Body Humanoid Control for Scene Interaction via Guided Diffusion

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Kalaria, Dvij, Harithas, Sudarshan S, Katara, Pushkal, Kwak, Sangkyung, Bhagat, Sarthak, Sastry, Shankar, Sridhar, Srinath, Vemprala, Sai, Kapoor, Ashish, Huang, Jonathan Chung-Kuan
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866908567880073216
author Kalaria, Dvij
Harithas, Sudarshan S
Katara, Pushkal
Kwak, Sangkyung
Bhagat, Sarthak
Sastry, Shankar
Sridhar, Srinath
Vemprala, Sai
Kapoor, Ashish
Huang, Jonathan Chung-Kuan
author_facet Kalaria, Dvij
Harithas, Sudarshan S
Katara, Pushkal
Kwak, Sangkyung
Bhagat, Sarthak
Sastry, Shankar
Sridhar, Srinath
Vemprala, Sai
Kapoor, Ashish
Huang, Jonathan Chung-Kuan
contents We introduce DreamControl, a novel methodology for learning autonomous whole-body humanoid skills. DreamControl leverages the strengths of diffusion models and Reinforcement Learning (RL): our core innovation is the use of a diffusion prior trained on human motion data, which subsequently guides an RL policy in simulation to complete specific tasks of interest (e.g., opening a drawer or picking up an object). We demonstrate that this human motion-informed prior allows RL to discover solutions unattainable by direct RL, and that diffusion models inherently promote natural looking motions, aiding in sim-to-real transfer. We validate DreamControl's effectiveness on a Unitree G1 robot across a diverse set of challenging tasks involving simultaneous lower and upper body control and object interaction. Project website at https://genrobo.github.io/DreamControl/
format Preprint
id arxiv_https___arxiv_org_abs_2509_14353
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle DreamControl: Human-Inspired Whole-Body Humanoid Control for Scene Interaction via Guided Diffusion
Kalaria, Dvij
Harithas, Sudarshan S
Katara, Pushkal
Kwak, Sangkyung
Bhagat, Sarthak
Sastry, Shankar
Sridhar, Srinath
Vemprala, Sai
Kapoor, Ashish
Huang, Jonathan Chung-Kuan
Robotics
Artificial Intelligence
Machine Learning
We introduce DreamControl, a novel methodology for learning autonomous whole-body humanoid skills. DreamControl leverages the strengths of diffusion models and Reinforcement Learning (RL): our core innovation is the use of a diffusion prior trained on human motion data, which subsequently guides an RL policy in simulation to complete specific tasks of interest (e.g., opening a drawer or picking up an object). We demonstrate that this human motion-informed prior allows RL to discover solutions unattainable by direct RL, and that diffusion models inherently promote natural looking motions, aiding in sim-to-real transfer. We validate DreamControl's effectiveness on a Unitree G1 robot across a diverse set of challenging tasks involving simultaneous lower and upper body control and object interaction. Project website at https://genrobo.github.io/DreamControl/
title DreamControl: Human-Inspired Whole-Body Humanoid Control for Scene Interaction via Guided Diffusion
topic Robotics
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2509.14353