Learning Versatile Humanoid Manipulation with Touch Dreaming

TL;DR

This paper introduces a multimodal Transformer-based system for humanoid robot manipulation that leverages touch, vision, and proprioception, achieving significant success in contact-rich tasks.

Contribution

It presents a novel touch dreaming approach with a Transformer model trained via behavioral cloning, enhancing dexterity and contact-awareness in humanoid robots.

Findings

HTD achieves 90.9% success rate across five tasks.

Latent-space tactile prediction outperforms raw tactile prediction.

The system enables versatile, high-dexterity manipulation in real-world scenarios.

Abstract

Humanoid robots promise general-purpose assistance, yet real-world humanoid loco-manipulation remains challenging because it requires whole-body stability, end-effector dexterity, and contact-aware interaction under frequent contact changes. In this work, we study dexterous, contact-rich humanoid loco-manipulation. We first develop an RL-based lower-body controller that serves as the stability backbone for whole-body execution during complex manipulation. Built on this controller, we develop a VR-based whole-body humanoid data collection system that integrates dexterous hands and tactile sensing for contact-rich manipulation. We then propose Humanoid Transformer with Touch Dreaming (HTD), a multimodal encoder--decoder Transformer that models touch as a core modality alongside multi-view vision and proprioception. HTD is trained in a single stage with behavioral cloning augmented by touch dreaming: in addition to predicting action chunks, the policy predicts future hand-joint forces and future tactile latents, with tactile-latent targets provided by an exponential moving average target encoder without requiring a separate tactile pretraining stage. This encourages the policy to learn contact-aware representations for dexterous manipulation. Across five real-world contact-rich tasks, HTD achieves a 90.9% relative improvement in average success rate over the stronger baseline. Ablation results further show that latent-space tactile prediction is more effective than raw tactile prediction, yielding a 30% relative gain in success rate. These results demonstrate that our touch-dreaming-enhanced learning system enables versatile, high-dexterity humanoid manipulation in the real world. More information and open-source materials are available at: humanoid-touch-dream.github.io.