Associative Skill Memory Models

Girgin, Hakan; Uğur, Emre

doi:10.1109/iros.2018.8593450

Cited by 10 publications

(7 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As an example, if the end-effector had carried a heavy object that significantly changed the measured forces when the robot was required to be compliant, the weight of the objects should have been used as the parameter to the model. In a previous study, 31 we showed the advantage of using the weight of the object as the parameter to the force-feedback model in an object pushing task. While we had used PHMMs to encode the force-feedback trajectory, the robot was provided only one trajectory to learn: the movement of the end-effector along a straight line while pushing an object.…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Compliant Parametric Dynamic Movement Primitives

Uğur¹,

Girgin²

2019

Robotica

Self Cite

View full text Add to dashboard Cite

SummaryIn this paper, we propose and implement an advanced manipulation framework that enables parametric learning of complex action trajectories along with their haptic feedback profiles. Our framework extends Dynamic Movement Primitives (DMPs) method with a new parametric nonlinear shaping function and a novel force-feedback coupling term. The nonlinear trajectories of the action control variables and the haptic feedback trajectories measured during execution are encoded with parametric temporal probabilistic models, namely parametric hidden Markov models (PHMMs). PHMMs enable autonomous segmentation of a taught skill based on the statistical information extracted from multiple demonstrations, and learning the relations between the model parameters and the properties extracted from the environment. Hidden states with high-variances in observation probabilities are interpreted as parts of the skill that could not be reliably learned and autonomously executed due to possibly uncertain or missing information about the environment. In those parts, our proposed force-feedback coupling term, which computes the deviation of the actual force feedback from the one predicted by the force-feedback PHMM, acts as a compliance term, enabling a human to scaffold the ongoing movement trajectory to accomplish the task. Our method is verified in a number of tasks including a real pick and place task that involves obstacles of different heights. Our robot, Baxter, successfully learned to generate the trajectory taking into the heights of the obstacles, move its end effector stiffly (and accurately) along the generated trajectory while passing through apertures, and allow human–robot collaboration in the autonomously detected segments of the motion, for example, when the gripper picks up the object whose position is not provided to the robot.

show abstract

Section: Resultsmentioning

confidence: 99%

“…We already studied such correcting actions in ref. [31] where our system exploited those models to correct the perturbed movements during executions with the aim of generalizing to novel configurations.…”

Section: Compliance In Parametric Dmpsmentioning

confidence: 99%

Compliant Parametric Dynamic Movement Primitives

Uğur¹,

Girgin²

2019

Robotica

Self Cite

View full text Add to dashboard Cite

show abstract

“…12,[19][20][21] Others have integrated learning meta and shape parameters, either hierarchically 9 or in parallel. [21][22][23] While RL Motion adaptation using manifold of task and movement primitive parameters 3 indeed facilitates acquiring new skills, its use for adaptation during runtime has several limitations. The continuous parameter space along with the high dimensionality of typical robotic motor-learning problems lead to long learning epochs.…”

Section: Related Workmentioning

confidence: 99%

“…Rueckert et al 28 suggest using hierarchical Bayesian models for estimating both meta and shape parameters for probabilistic movement primitives. Ugur and Girgin 20,21 suggest learning parametric hidden Markov models from multiple demonstrations for encoding relations between shape parameters and environment properties. Probabilistic movement primitives facilitate encoding optimal behaviors in stochastic systems, yet when deterministic system behavior (i.e., completely predictable operation) is required, they are less suitable.…”

Section: Related Workmentioning

confidence: 99%

Motion Adaptation Based on Learning the Manifold of Task and Dynamic Movement Primitive Parameters

2020

View full text Add to dashboard Cite

SUMMARY Dynamic movement primitives (DMP) are motion building blocks suitable for real-world tasks. We suggest a methodology for learning the manifold of task and DMP parameters, which facilitates runtime adaptation to changes in task requirements while ensuring predictable and robust performance. For efficient learning, the parameter space is analyzed using principal component analysis and locally linear embedding. Two manifold learning methods: kernel estimation and deep neural networks, are investigated for a ball throwing task in simulation and in a physical environment. Low runtime estimation errors are obtained for both learning methods, with an advantage to kernel estimation when data sets are small.

show abstract

“…Memorized force and tactile profiles have already been successfully utilized in modulating learned DMPs in difficult manipulation tasks that contain high degrees of noise in perception such as grasping and in-hand manipulation of objects from incorrect positions or flipping boxes using chopsticks [17]. HMMs were used to learn multi-modal models from temperature, pressure and fingertip information for exploratory object classification tasks [7]; and PHMMs were used to learn haptic feedback trajectory models to adapt actions in response to external perturbations [10,21]. Deep networks [8] were used to learn multi-modal models from different sensory information such as temperature, pressure, fingertip, contacts, proprioception, and speech; however these models were used only to categorize the sensory data without any effect on action execution.…”

Section: Related Workmentioning

confidence: 99%

Conditional Neural Movement Primitives

Seker

Imre

Piater

et al. 2019

Robotics: Science and Systems XV

Self Cite

View full text Add to dashboard Cite

Conditional Neural Movement Primitives (CNMPs) is a learning from demonstration framework that is designed as a robotic movement learning and generation system built on top of a recent deep neural architecture, namely Conditional Neural Processes (CNPs). Based on CNPs, CNMPs extract the prior knowledge directly from the training data by sampling observations from it, and uses it to predict a conditional distribution over any other target points. CNMPs specifically learns complex temporal multi-modal sensorimotor relations in connection with external parameters and goals; produces movement trajectories in joint or task space; and executes these trajectories through a high-level feedback control loop. Conditioned with an external goal that is encoded in the sensorimotor space of the robot, predicted sensorimotor trajectory that is expected to be observed during the successful execution of the task is generated by the CNMP, and the corresponding motor commands are executed. In order to detect and react to unexpected events during action execution, CNMP is further conditioned with the actual sensor readings in each time-step. Through simulations and real robot experiments, we showed that CNMPs can learn the nonlinear relations between low-dimensional parameter spaces and complex movement trajectories from few demonstrations; and they can also model the associations between high-dimensional sensorimotor spaces and complex motions using large number of demonstrations. The experiments further showed that even the task parameters were not explicitly provided to the system, the robot could learn their influence by associating the learned sensorimotor representations with the movement trajectories. The robot, for example, learned the influence of object weights and shapes through exploiting its sensorimotor space that includes proprioception and force measurements; and be able to change the movement trajectory on the fly when one of these factors were changed through external intervention.

show abstract

Associative Skill Memory Models

Cited by 10 publications

References 20 publications

Compliant Parametric Dynamic Movement Primitives

Compliant Parametric Dynamic Movement Primitives

Motion Adaptation Based on Learning the Manifold of Task and Dynamic Movement Primitive Parameters

Conditional Neural Movement Primitives

Contact Info

Product

Resources

About