Expert-Play Contouring Control: Faster-than-Demonstration Planning from Slow Expert and Fast Play

Seunghoon Cho; Wonsuhk Jung; Sundhar Vinodh Sangeetha; Shreyas Kousik

cs.ROSep 19, 2026

Expert-Play Contouring Control: Faster-than-Demonstration Planning from Slow Expert and Fast Play

Authors: Seunghoon Cho, Wonsuhk Jung, Sundhar Vinodh Sangeetha, Shreyas Kousik

Organizations: Seoul National University, Seoul, South Korea · Georgia Institute of Technology, Atlanta, GA, USA

Abstract

Expert demonstrations often specify what a robot should do, but not how fast it can do it. Imitation Learning (IL) inherits demonstration timing, while directly accelerating the learned motion can fail when faster execution changes the robot-object dynamics. We study faster-than-demonstration execution as a dynamics-aware control problem and introduce Expert-Play Contouring Control (EPCC), which combines slow expert demonstrations with fast, non-expert play. Expert demonstrations train a latent trajectory generator whose predictions are reparameterized into a time-independent contour of successful task progression, while play trains a world model (WM) of fast-action outcomes. At deployment, our proposed planner uses the WM to optimize actions that makes maximize progress along the expert-derived contour while penalizing deviation from the intended task evolution. Averaged across three visuomotor manipulation tasks, EPCC achieves a $2.0\times$ the throughput of the IL baseline, including $2.2\times$ that of the throughput of the strongest acceleration baseline on a task with interaction-sensitive object dynamics. Our analysis shows that the gains concentrate where faster execution changes robot-object evolution. Together, our results highlight a simple yet effective principle: demonstrations provide task intent, while play data provides the dynamic coverage needed to execute that intent faster.

Figures & tables

Fig. 2 : Overview of EPCC . Training (left): play data $\mathcal{D}_{\mathrm{\textnormal{p}}}$ trains a latent world model, consisting of an encoder $h_{\phi}$ and an action-conditioned predictor $f_{\theta}$ , with the next-latent prediction objective $\mathcal{L}_{\mathrm{\textnormal{WM}}}$ ( Section IV-A ). Expert demonstrations $\mathcal{D}_{\mathrm{\textnormal{e}}}$ are encoded with the frozen $h_{\phi}$ and used to train a latent trajectory generator $G_{\psi}$ ; the path reparameterization $\mathcal{P}$ removes the demonstrations’ timing, yielding a time-independent contour $\bar{z}_{t}(s)$ , $s\in[0,1]$ ( Section IV-B ). Inference (right): from the current observation, the contour generator $C_{\psi}$ produces the reference contour, and the controller jointly optimizes the action sequence $A_{t}$ and progress increments $V_{t}$ by minimizing the contouring error $e^{c}_{t+k|t}$ between the world-model rollout $\hat{z}_{t+k|t}$ and $\bar{z}_{t}(s_{t+k|t})$ , maximizing progress $\Delta s$ along the contour, and regularizing the actions ( Section IV-C , Equation 25 ). The first ${T_{a}}$ actions are executed and both the contour and the plan are recomputed from the next observation.

	PushT			Cube			Slippery-PushT
	SR	TTC	TP	SR	TTC	TP	SR	TTC	TP
	(%)	(s)	(1/s)	(%)	(s)	(1/s)	(%)	(s)	(1/s)
DP	88	24.3	.033	96	6.8	.137	68	28.8	.021
SkiP	56	20.2	.019	80	4.9	.146	24	23.6	-.003
SAIL	66	19.8	.033	96	3.8	.240	46	18.5	.016
SuP	92	12.1	.070	100	3.2	.321	48	19.8	.017

TABLE I : Main benchmark. Metrics are defined in Section V-A ; leading zeros are omitted from TP. Higher is better for SR and TP, lower for TTC. TTC is the median over successful episodes: ‘–’ marks a cell with no successful episode. Bold (resp. underline) marks the best (resp. second best) entry in each metric column.

Panel of Figure 3 — Fig. 3 : When is dynamics-aware acceleration helpful? (a) SR, time to completion, and TP across the Slippery-PushT with varying block’s coasting time $\tau$ . (b) Speed-up ratio $\alpha$ during contact and free motion, median over each method’s successful episodes. EPCC gains most as interaction dynamics become more consequential and during contact-sensitive phases.

	PushT			Cube			Slippery-PushT
	SR	TTC	TP	SR	TTC	TP	SR	TTC	TP
	(%)	(s)	(1/s)	(%)	(s)	(1/s)	(%)	(s)	(1/s)
Select	92	12.6	.069	96	3.5	.270	16	24.5	-.005
EPCC-WS	94	10.0	.092	78	2.6	.281	60	20.2	.030
EPCC	78	10.4	.073	84	2.5	.317	62	15.8	.033

TABLE II : Play-supported action search vs. temporal scaling. Select is restricted to temporally rescaled actions from the frozen policy, while EPCC-WS and EPCC can optimize actions beyond this candidate set using the action priors in play data. Success rate (SR), time-to-completion (TTC), and throughput (TP) are reported over

50

episodes.

Fig. 4 : What play data enables acceleration? EPCC performance with equal-budget play datasets collected at different paces. Dashed lines denote Expert-Only ; Play data at $2.6$ - $3.5\times$ demonstration pace achieves highest throughput.

Fig. 5 : Scaling with play data. Throughput of EPCC across tasks as the amount of play data used to train WM increases.

Fig. 6 : Real-world setup. A Kinova Gen3 arm uses a ZED scene camera to push a T-shaped object into the target region on a low-friction surface, where faster execution makes object motion more sensitive to contact dynamics.

Appendix figures & tables5 assets

Supplementary material from the paper’s appendix.

Appendix

	PushT	Slippery-PushT	Cube
Execution Horizon	$3$	$1$	$3$
CEM candidates	$300$	$300$	$1000$
Progress weight $w_{p}$	$4.21$	$9.26$	$6.01$
Progress cap $\tilde{\nu}_{\max}$	$4.0$	$2.7$	$4.0$
Smoothness $r_{\Delta}$	$0$	$1$	$1$

Appendix

TABLE III : Main-benchmark controller settings.

Fig. 7 : Additional play-data scaling results. Success rate (top) and median completion time among successes (bottom) across dataset sizes; the corresponding throughput results appear in Figure 5 . TTC is omitted for fewer than five successful episodes.

	PushT	Slippery-PushT	Cube
Expert-Only	— / $4.0$	— / $2.2$	— / $2.8$
Play 1	$1.85$ / $2.0$	$1.44$ / $1.1$	$1.06$ / $0.84$
Play 2	$3.51^{\dagger}$ / $4.0$	$2.47^{\dagger}$ / $2.2$	$2.65$ / $2.1$
Play 3	$4.15$ / $6.0$	$3.28$ / $3.3$	$3.54^{\dagger}$ / $2.8$
Play 4	$5.30$ / $10.0$	$4.21$ / $5.5$	$4.22$ / $3.4$

Appendix

TABLE IV : Play collection pace (relative to expert demonstrations) and corresponding controller progress cap for Figure 4 . Entries are pace /

\tilde{\nu}_{\max}

; † denotes the nominal

1\times

play condition.

Panel of Figure 11 — Fig. 8 : Simulation tasks. PushT and Slippery-PushT share the same pushing geometry but differ in block damping; Cube requires grasping and placing an object.

Task	$O$	$B$	$D$	Total	Episodes
PushT	$62.7$	$8.0$	$6.9$	$77.6$	$21{,}393$
Slippery-PushT	$30.5$	$16.4$	—	$46.8$	$7{,}880$
Cube	$27.9$	$18.9$	$10.7$	$57.5$	$19{,}111$

Appendix

TABLE V : Main-benchmark play mixtures (hours) used in Table I .

Robot Motion Planning

Robot motion planning finds safe, feasible paths and trajectories for robots moving through constrained, changing environments.

See all 271 papers

Retiming Robot Demonstrations with a Learned Speed Prior
Oct 23, 2026Seungyong Moon, Deokjae Lee, Hansang Lee +7Robot Policy Learning Learning from Demonstration
Seoul National University
EgoSpeedUp: Transferring Human Manipulation Tempo to Robot Policies
Sep 24, 2026Hanbit Oh, Yukiyasu Domae, Takuma YagiRobot Policy Learning Robotic Manipulation
Artificial Intelligence Research Center, National Institute of Advanced Industrial Science and Technology (AIST), Japan
AutoSpeed: Annotation-Free Stage-Adaptive Motion Speed Learning for Robot Manipulation
Jul 1, 2026Qingda Hu, Ziheng Qiu, Jieru Zhao +2Robot Skill Learning Visuomotor Policy Learning
College of Intelligent Robotics and Advanced Manufacturing, Fudan University, Shanghai, China · School of Computing, Shanghai Jiao Tong University, Shanghai, China

Oct 23, 2026J/K move · Enter open · S save

Retiming Robot Demonstrations with a Learned Speed Prior

Seungyong Moon, Deokjae Lee, Hansang Lee, Munkhtselmeg Battsengel+6

Seoul National University

Robot Policy Learning Learning from Demonstration Robot Imitation Learning

Open paper AskPDF arXiv

The pace of a robot demonstration is often determined by the teleoperator rather than by the task itself. As a result, the same motion can appear at widely different speeds across demonstrations, and policies trained on such data may inherit operator-specific timing along with task behavior. We propose to learn a speed prior directly from large-scale demonstration data. Given a task instruction, a motion path and the images along it, it predicts a distribution over execution speed along the path, from which a chosen quantile defines a consistent clock for retiming each demonstration. Existing methods derive the acceleration signal from the very data they accelerate, so each new task costs another fit, annotation, or pair of predefined rates. A single trained speed prior instead covers every task in a corpus. We fit a speed prior on both robot teleoperation data and handheld demonstrations in which a human arm, rather than a robot arm, moves the gripper, and use it to retime robot demonstrations to the handheld pace before fine-tuning a vision-language-action policy. For execution, we pair the policy with a real-time optimal-control layer that tracks predicted action chunks under robot constraints using the predicted pace. On 120 RoboLab tasks, policies fine-tuned on our learned clocks succeed more often and move with up to half the jerk compared with policies fine-tuned on the original demonstration timing and uniformly accelerated to the same speed. On a real robot, our policy stacks cups in 3.50s, 1.72

\times

faster than the default policy's 6.02s while succeeding as often, in 29 of 30 trials. Videos and further material are available at https://mllab.snu.ac.kr/projects/learned_speed_prior.

Abstract

Robot Motion Planning

Explore similar work

Retiming Robot Demonstrations with a Learned Speed Prior

EgoSpeedUp: Transferring Human Manipulation Tempo to Robot Policies

AutoSpeed: Annotation-Free Stage-Adaptive Motion Speed Learning for Robot Manipulation