Dressing in Motion: A Human Motion-Aware Diffusion Policy for Robot-Assisted Dressing

Haoxiang Sun; Fangyuan Wang; Songhao Huang; Justina Y. W. Liu; Jihong Zhu; Peng Zhou; David Navarro-Alarcon

cs.ROSep 4, 2026

Dressing in Motion: A Human Motion-Aware Diffusion Policy for Robot-Assisted Dressing

Authors: Haoxiang Sun, Fangyuan Wang, Songhao Huang, Justina Y. W. Liu, Jihong Zhu, Peng Zhou, David Navarro-Alarcon

Organizations: Department of Mechanical Engineering, The Hong Kong Polytechnic University, Hong Kong SAR, China. · School of Nursing, The Hong Kong Polytechnic University, Hong Kong SAR, China. · School of Physics, Engineering and Technology, University of York, UK. · School of Advanced Engineering, Great Bay University, Dongguan, Guangdong, China.

Abstract

Robotic dressing assistance is a promising solution for supporting older adults with physical impairments in daily living. However, dressing under human motion remains challenging, as complex garment--human contact and occlusions make it difficult to generate actions aligned with arm movements. In this letter, we propose a visuomotor policy that learns dressing skills from static expert demonstrations and generalizes to dynamic user-motion scenarios. A diffusion policy tailored to garment--human interaction geometry learns from partially observed point clouds with varied arm postures. We then introduce an object-centric representation based on PDE diffusion to capture the axial distribution of the arm. By sampling motion-relevant regions and registering them across consecutive observations, the proposed method approximates arm motion and reactively adapts the executed trajectory. We evaluate our method in simulation and a real-world human study involving nine participants, three garment types, and six arm-motion patterns. Results show that our method outperforms baselines in dressing progress, freedom of movement, and user comfort. The project website is https://anonymous.4open.science/w/dressing-in-motion.

Figures & tables

Fig. 1: Robot-assisted dressing under human arm motion. The robot advances the garment along the arm axis while adapting to time-varying arm geometry.

Fig. 2: Overview of the proposed method. Top left : Visual encoder extracts point-cloud features and fuses them with the robot state to obtain the observation $\mathbf{o}_{t}$ . Top right : A diffusion policy denoise Gaussian noise to predict an action chunk. Bottom : Diffused scalar field identifies the region $\mathcal{R}_{t+1}$ , which is registered to the reference region $\mathcal{R}_{t}$ to estimate arm motion. The estimated transformation is applied to generate the adapted execution trajectory.

Fig. 3: Diffused scalar-field construction and ROI sampling. Surface diffusion assigns each arm point a scalar value $u$ encoding its geodesic position along the arm axis. The uncovered arm region $\mathcal{U}$ is extracted, from which the region of interest $\mathcal{R}$ is sampled.

Fig. 4: Examples of dressing process with different garment types and human models in the simulation environment.

Panel of Figure 5 — Fig. 5: Simulation performance under different arm-motion speeds. For each benchmark and speed profile, results are computed over 70 randomized trials. Top : dressing ratio. Bottom : sleeve insertion success rate.

Fig. 6: Error bar plots of the diffused scalar field and point-cloud registration under different diffusion times $\tau$ . The plots report the mean RMSE over 50 independent randomized trials, with error bars indicating standard deviations.

Fig. 7: Top left : Real-world experimental setup. Top right : garments with different sleeve lengths. Bottom : motion patterns performed by the participants.

Fig. 8: Snapshots of real-world dressing trials using the proposed method with different participants. Participants select varied initial arm configurations, including arm orientations and shoulder and elbow joint angles. The policy first aligns the garment opening with the hand and then adaptively updates the execution trajectory in response to subsequent arm motion.

Fig. 9: Dressing outcomes of the real-world human study. Left : dressing ratio. Right : sleeve insertion success rate.

Method	Vest	Shirt	Windbreaker
Ours	0.92	0.89	0.83
Ours w/o TA	0.93	0.92	0.50
DP3	0.79	0.58	0.67
DP-image	0.45	0.33	0.25
Diff-MPC	0.14	0.30	0.27
BC-LSTM	0.58	0.50	0.46

TABLE I: Dressing Ratio across Garments

Method	Up-down	Side2side	Back-forth	Wave	Phone	Object
Ours	1.00	1.00	1.00	1.00	0.77	0.50
Ours w/o TA	0.83	0.83	0.83	0.83	0.76	0.60
DP3	0.50	0.83	0.50	1.00	0.75	0.50
DP-image	0.31	0.17	0.50	0.50	0.33	0.25
Diff-MPC	0.29	0.42	0.20	0.20	0.12	0.20
BC-LSTM	1.00	0.17	0.50	0.92	0.49	0.00

TABLE II: Dressing Ratio across Motion Patterns

Fig. 10: Subjective rating results from the real-world human study. Top : freedom of movement. Bottom : comfort. Left : box plots of the rating results. Right : full distributions of participant ratings.

Fig. 11: Results of dressing with different arm motion amplitudes and speeds. Left : horizontal motion. Right : vertical motion.

Explore similar work

Wearing A Coat: Dual-Arm Robot-Assisted Dressing with Differentiable Clothing Simulation
Jul 13, 2026Yiming Liu, Lijun Han, Hesheng WangContact-Rich Robotic Manipulation Assistive Robotics
Department of Automation, Shanghai Jiao Tong University, Shanghai 200240, China
Learning Robot Policies for Whole-Body Mobility Assistance via Dynamic Human Motion Generation
Oct 13, 2026Shikun Ban, Shohei Nagai, Qingyang Song +4Assistive Robotics Robot Imitation Learning
The Robotics Institute, School of Computer Science, Carnegie Mellon University Pittsburgh, USA · Intuitive Autonomy San Jose, USA
DROM: A Language-Guided Diffusion Framework for Multi-Skill Robotic Manipulation
Sep 29, 2026Vincenzo Pomponi, Rocco Felici, Paolo Franceschi +5Long-Horizon Robotic Manipulation Robot Skill Learning
Institute of Systems and Technologies for Sustainable Production (ISTePS), Department of Innovative Technologies, University of Applied Science and Arts of Southern Switzerland (SUPSI), Via la Santa 1, Lugano, CH-6900, Ticino, Switzerland. · Istituto Dalle Molle di studi sull’intelligenza artificiale (IDSIA), Department of Innovative Technologies, University of Applied Science and Arts of Southern Switzerland (SUPSI), Via la Santa 1, Lugano, CH-6900, Ticino, Switzerland. · Mechanical Department, Politecnico di Milano (PoliMi), Via Giuseppe Candiani, Milan, 20155, Italy. +1

Jul 13, 2026J/K move · Enter open · S save

Wearing A Coat: Dual-Arm Robot-Assisted Dressing with Differentiable Clothing Simulation

Yiming Liu, Lijun Han, Hesheng Wang

Department of Automation, Shanghai Jiao Tong University, Shanghai 200240, China

Contact-Rich Robotic Manipulation Assistive Robotics Differentiable Physics

Open paper AskPDF arXiv

The development of assistive robots for dressing tasks serves to augment human convenience and improve the quality of life for individuals with physical impairments. However, due to the intricate contact interactions between garments and the human limbs during dressing, most robot-assisted dressing algorithms treat clothing as an assembly of discrete segments, thereby struggling to manage the partial worn garments under contact constraints. To overcome this challenge, we propose a novel robotic dressing control algorithm that integrates realtime differentiable clothing simulation. The simulation algorithm employs explicit iterative scheme with intentionally introduced higher-order perturbations to enhance computational efficiency while maintaining stability under large time-step conditions. Through simulation, we resolve the garment state under contact constraints, which then enables a multi-phase control strategy for successful coat dressing assistance. To further improve real-time performance, we introduce a constrained local model along with its corresponding optimization solver, permitting high-frequency local compensation for the differentiable simulation based global controller. Finally, we experimentally validate our approach through both simulated and physical dressing scenarios, conclusively demonstrating its feasibility and efficacy

Abstract

Explore similar work

Wearing A Coat: Dual-Arm Robot-Assisted Dressing with Differentiable Clothing Simulation

Learning Robot Policies for Whole-Body Mobility Assistance via Dynamic Human Motion Generation

DROM: A Language-Guided Diffusion Framework for Multi-Skill Robotic Manipulation