cs.ROMar 11, 2026

ContactExplorer: Contact Coverage-Guided Exploration for General-Purpose Dexterous Manipulation

Authors: Zixuan Liu, Ruoyi Qiao, Chenrui Tie, Xuanwei Liu, Yunfan Lou, Chongkai Gao, Zhixuan Xu, Lin Shao

Organizations: School of Computing, National University of Singapore · RoboScience

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

Reinforcement learning explores effectively in domains such as Atari games, navigation, and locomotion, where novelty over states or dynamics is a sufficient signal. In contrast, dexterous manipulation requires rich physical hand--object interactions, but existing methods often suffer from unstable contact-based novelty signals, inefficient distance novelty signals, or reliance on task-specific priors. We propose ContactExplorer, a general exploration method for dexterous manipulation tasks. ContactExplorer represents contact as the intersection between object surface points and hand keypoints, encouraging dexterous hands to discover diverse and novel contact patterns, namely which fingers contact which object regions. It maintains a contact counter conditioned on discretized object states obtained via learned hash codes. This counter is leveraged in two complementary ways: (1) a count-based contact coverage reward that promotes exploration of novel contact patterns, and (2) an energy-based reaching reward that guides the agent toward under-explored contact regions. We evaluate ContactExplorer on seven contact-rich manipulation tasks and five dexterous hand embodiments. Experimental results show that ContactExplorer substantially improves sample efficiency and success rates over existing exploration methods, that it reduces the need for task-specific priors, and that it remains effective across hand embodiments and transfers to the real world. Project page is https://contact-explorer.github.io.

Figures & tables

Fig. 1: Contact Exploration Strategies. (A) Existing strategies discover contact inefficiently, perturbing only nearby contacts, exploring without requiring contact, or relying on forces that are unreliable under discontinuous contacts. (B) ContactExplorer tracks contact coverage over object surface regions and guides the fingers toward under-covered contacts, yielding effective interaction and faster convergence.
Fig. 2: Hand Keypoint Representation. Sparse surface keypoints on each finger.
Fig. 3: Overview of ContactExplorer. Our contact coverage-guided exploration framework consists of three main components: (1) a learned state hashing module that clusters continuous object states (Sec. III-C ); (2) a contact coverage counter that tracks state-conditioned finger-region interactions (Sec. III-D ); and (3) a structured exploration reward mechanism, decomposed into contact coverage (Sec. IV-A ) and energy-based reaching terms (Sec. IV-B ). In the visualization, the current and goal object states are depicted as distinctly colored point clouds.
Fig. 4: Learning curves on seven dexterous manipulation tasks. Each panel plots success rate against environment steps for ContactExplorer and four baselines ( TR , LHCC , HaC , RND-Dist ), averaged over five random seeds; shaded regions denote ±1\pm 1 standard deviation across seeds. ContactExplorer attains higher final success rates and faster convergence, and is the only method to make progress on the hard-exploration Constrained Object Retrieval task, where all baselines fail.
Fig. 5: Task overview. Seven contact-rich tasks, each demanding a distinct interaction pattern. (a) Cluttered Object Singulation : extract a target from a densely packed row. (b) Constrained Object Retrieval : retrieve an object from a walled space. (c) In-Hand Reorientation : rotate an object to a goal orientation. (d) Bimanual Object Opening : two hands open an articulated object. (e) Bimanual Board Lifting : two hands lift an oversized board. (f) Grasping : grasp and lift a target. (g) Tiled Object Retrieval : retrieve a cube from a tightly tiled bin.
BudgetMethodTaskAvg.
SingulationConstrained RetrievalIn-Hand ReorientationBimanual OpeningBoard LiftingGraspingTiled Retrieval
40%TR72{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 34}}0{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 0}}64{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 16}}74{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 25}}27{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 37}}52{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 25}}39{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 7}}4747
LHCC [ 12 , 43 ]34{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 26}}0{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 0}}64{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 14}}67{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 25}}18{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 37}}44{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 25}}42{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 14}}3838
HaC [ 14 ]14{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 19}}0{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 0}}64{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 13}}51{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 38}}38{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 47}}29{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 23}}41{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 6}}3434
RND-Dist [ 13 ]67{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 33}}0{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 0}}66{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 13}}69{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 26}}19{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 39}}27{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 31}}43{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 7}}4242
Ours\mathbf{94}{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 3}}\mathbf{67}{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 34}}\mathbf{76}{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 14}}\mathbf{85}{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 9}}\mathbf{75}{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 16}}\mathbf{72}{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 8}}\mathbf{57}{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 11}}75\mathbf{75}
TABLE I: Main results across all tasks under full and 40% training budgets. Per-task SR in %, mean ±\pm std over five seeds; best in bold. Avg. is the unweighted mean of the seven per-task means.
Fig. 6: Task-Specific Initialization Prior. Default and contact-aware initialization in Constrained Object Retrieval .
Method40% budgetFull budget
TR0{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 0}}0{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 0}}
TR-ContactInit25{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 32}}33{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 40}}
Ours\mathbf{67}{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 34}}\mathbf{88}{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 6}}
TABLE II: Comparison with Task-Specific Prior. SR in %, mean ±\pm std over five seeds; best in bold.
Reward terms40% budgetFull budget
Sing.Constrained Retri.Sing.Constrained Retri.
Energy only88{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 4}}12{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 15}}94{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 2}}36{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 44}}
Contact only73{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 23}}31{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 38}}89{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 7}}36{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 44}}
Ours\mathbf{94}{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 3}}\mathbf{67}{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 34}}\mathbf{96}{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 1}}\mathbf{88}{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 6}}
TABLE III: Reward Ablation Studies. Sing. and Constrained Retri. denote Cluttered Object Singulation and Constrained Object Retrieval, respectively. SR in %, mean ±\pm std over five seeds; best in bold.
Fig. 7: Box Push diagnostic. Learned state clusters keep the contact statistics of the two initial configurations separate.
Fig. 9: Open-Loop Trajectory Replay. Replaying privileged teacher actions open-loop on the real robot produces consistent 90 ​ ° zz -axis rotations.
Fig. 10: Real-World Policy Execution. A representative rollout (left to right) of the learned policy performing Cluttered Object Singulation .
MethodExposureTransport
TR11/30 (36.7%)11/30~(36.7\%)1/30 (3.3%)1/30~(3.3\%)
Ours23/30 (76.7%)\mathbf{23/30~(76.7\%)}10/30 (33.3%)\mathbf{10/30~(33.3\%)}
TABLE IV: Real-World Cluttered Object Singulation Results. Successes over 30 trials; Transport additionally requires moving the exposed object to the goal.
ParameterValue
Parallel environments2,048
Learning-rate scheduleKL-adaptive
PPO epochs per update5
Minibatches per epoch4
Discount factor γ\gamma0.99
GAE parameter λGAE\lambda_{\mathrm{GAE}}0.95
TABLE V: PPO Hyperparameters.
ParameterIsaac Gymmjlab
ManipulationReorientationReorientationGrasping
Contact coefficient α\alpha200240128128
Reaching coefficient β\beta1.282.560.50.5
Hash length HH5855
Surface points MM512 *512512512
Surface regions KK32326464
TABLE VI: ContactExplorer Parameters. Manipulation includes single-arm and bimanual tasks. AE denotes the object-state autoencoder. ∗ Board lifting uses 1,024 surface points; object opening uses up to 1,024 per articulated part.
α\alphaβ\betaSuccess rate (%)
  40% budgetFull budget
50500.320.3214{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 28}}17{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 35}}
1001000.640.6430{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 37}}52{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 42}}
2002001.281.2867{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 34}}88{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 6}}
4004002.562.56\mathbf{69}{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 16}}\mathbf{89}{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 4}}
8008005.125.1240{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 35}}56{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 45}}
TABLE VII: Exploration Reward Scales. Constrained Object Retrieval SR in %, mean ±\pm std over five seeds. Best mean per column in bold; the default is shaded.
HHλ\lambdaSuccess rate (%)
  40% budgetFull budget
 0.50.568{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 25}}90{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 2}}
 1.01.063{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 32}}85{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 6}}
442.02.049{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 40}}52{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 43}}
 0.50.564{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 33}}73{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 36}}
 1.01.067{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 34}}88{\scriptstyle\,{\color[rgb]{0.55,0.55,0.55}\pm 6}}
TABLE VIII: Object-State Clustering Parameters. Constrained Object Retrieval SR in %, mean ±\pm std over five seeds. Best mean per column in bold; the default is shaded.
Fig. 12: Simulation and Real Point Clouds. Simulated points are blue and real observations are red.
ObservationDimension
Pt−1:tmask\mathbf{P}_{t-1:t}^{\mathrm{mask}}2×1024×42\times 1024\times 4
qt−1:tarm\mathbf{q}_{t-1:t}^{\mathrm{arm}}2×72\times 7
qt−1:thand\mathbf{q}_{t-1:t}^{\mathrm{hand}}2×162\times 16
TABLE IX: Student Observations. Two-step histories of the scene point cloud with a binary target-mask channel and robot joint positions.
Fig. 13: Real-World Failure Cases. Manipulating a non-target object (left) and failing to stabilize the target after singulation (right).

Explore similar work

CardsList
Sep 27, 2026cs.RO

DexTaG: Tactile-as-Guidance in Reinforcement Learning for Dexterous Manipulation

Glove-based motion capture is emerging as a scalable approach to collecting dexterous-hand demonstration data. However, due to the kinematic gap between the human and robot hand, the recorded human motions cannot be executed directly on the robot, especially for contact-rich tool-use tasks involving in-hand reorientation. Prior work bridges this gap in simulation through reinforcement learning (RL) or trajectory optimization, but the human contact pattern is hard to preserve under such formulations, often producing unnatural manipulation and unstable functional grasps. These methods also train a separate policy or solve a separate optimization for each reference trajectory, which is inefficient. To solve these problems, we propose DexTaG, a tactile-guided RL framework for dexterous manipulation. During training, tactile signals captured by the glove guide policy search toward the measured human contact pattern, reducing reliance on precise reference geometry for contact supervision. To improve efficiency, we train a single generalizable retargeter jointly on all training trajectories of the same object. The retargeter is further distilled into a tactile-free student controller conditioned on the target object trajectory for real-world deployment. On marker-pen and hammer manipulation tasks, DexTaG learns natural, contact-rich behaviors that baselines with distance-based contact heuristics fail to learn, generalizes to held-out trajectories of the same object and task, and outperforms single-trajectory baselines on OakInk2.

Han Yang, Yian Wang, Yunlong Song +2

University of Massachusetts Amherst. · Genesis AI.

Contact-Rich Dexterous ManipulationDexterous ManipulationTactileContact-Rich Manipulation
AskPDF
Jun 23, 2026cs.RO

NoContactNoWorries: Estimating Contact through Vision and Proprioception for In-Hand Dexterous Manipulation

Perceiving physical contact is fundamental to dexterous manipulation. While robots often rely on dedicated hardware tactile sensors, humans exhibit a remarkable ability to infer contact by integrating visual information with an innate sense of their body's pose and movement. Inspired by this embodied perceptual skill, we investigate whether a robot can learn to infer contact from vision, an approach that also offers a scalable alternative to tactile hardware specifically for binary contact estimation, which faces practical challenges in cost, fragility, and integration. We present NoContactNoWorries, a transformer-based multimodal framework that fuses RGB-D vision with the robot's proprioception to infer binary contact states as a pseudo-tactile signal for hand-object interactions. We validate by training a single contact prediction model on multiple objects and show that the inferred contact signal supports downstream reinforcement learning agents for in-hand object reorientation, generalizing to novel objects. Experiments in both simulation and on a real-world robot validate our approach, highlighting the feasibility of inferring contact from vision and proprioception. Project Page: https://soham2560.github.io/no-contact-no-worries/

Soham Patil, Avirup Das, Sourabh Bhosale +1

Robotics Research Center (RRC), International Institute of Information Technology (IIIT), Hyderabad, India · Department of Computer Science, The University of Manchester

Contact-Rich Dexterous ManipulationTactileDexterous ManipulationProprioceptionRobotic Systems
AskPDF
Aug 1, 2026cs.RO

DexMani: Human-Derived Manipulability Guidance for Dexterous Rotation

Dexterous object rotation is a sequential contact problem: each support, release, and re-contact decision must both produce the desired object motion, and prepare the hand configuration for continued rotation. Existing reinforcement learning methods discover such movement patterns through trial and error on specific robotic hand embodiments, without explicitly accounting for how each contact transition affects the hand's ability to sustain object rotation in subsequent steps. We introduce DexMani, a framework that transfers human demonstrations as contact-conditioned manipulability evolution. This prior captures how successful human contact transitions reshape the object-rotation directions available to the hand. DexMani then learns this manipulability evolution and uses it to guide downstream reinforcement learning, enabling rotation skills to be acquired across robot embodiments with distinct kinematics and active-contact configurations. Across the Shadow Hand, Allegro Hand, and XHand, DexMani achieves the highest success rates in every evaluated setting for both seen and unseen objects. DexMani reaches an average success rate of 57.5% on LEAP Hand, outperforming other baselines and producing smoother rotatory motions. Project site: https://dexmani.github.io

Xiaoyang Chen, Shengcheng Luo, Haoran Guo +4

Shanghai Jiao Tong University · ShanghaiTech University · Beijing Institute for General Artificial Intelligence (BIGAI) +1

Contact-Rich Dexterous ManipulationOffline Reinforcement Learning
AskPDF