Small-variance asymptotics for non-parametric online robot learning

Kumar, T.M. Kiran; CalinonSylvain,

doi:10.1177/0278364918816374

Cited by 15 publications

(17 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We present latent space formulations of our approach to exploit the task structure using: 1) mixture of factor analyzers decomposition of the covariance matrix [15], 2) semi-tied covariance matrices of the mixture model [25], and 3) Bayesian non-parametric formulation of the model with Hierarchical Dirichlet process (HDP) for online learning under small variance asymptotics [26]. The paper unifies and extends our previous work on encoding manipulation skills in a task-adaptive manner [24,25,26]. Our objective is to reduce the number of demonstrations required for learning a new task, while ensuring effective generalization in new environmental situations.…”

Section: Introductionmentioning

confidence: 99%

Generalizing Robot Imitation Learning with Invariant Hidden Semi-Markov Models

Tanwani

Thananjeyan

Laskey

et al. 2019

EasyChair Preprints

Self Cite

View full text Add to dashboard Cite

Generalizing manipulation skills to new situations requires extracting invariant patterns from demonstrations. For example, the robot needs to understand the demonstrations at a higher level while being invariant to the appearance of the objects, geometric aspects of objects such as its position, size, orientation and viewpoint of the observer in the demonstrations. In this paper, we propose an algorithm that learns a joint probability density function of the demonstrations with invariant formulations of hidden semi-Markov models to extract invariant segments (also called sub-goals or options), and smoothly follow the generated sequence of states with a linear quadratic tracking controller. The algorithm takes as input the demonstrations observed with respect to different coordinate systems describing virtual landmarks or objects of interest, and adapts the segments according to the environmental changes in a systematic manner. We present variants of this algorithm in latent space with low-rank covariance decompositions, semi-tied covariances, and non-parametric online estimation of model parameters under small variance asymptotics; yielding considerably low sample and model complexity for acquiring new manipulation skills. The algorithm allows a Baxter robot to learn a pick-and-place task while avoiding a movable obstacle based on only 4 kinesthetic demonstrations.

show abstract

Section: Introductionmentioning

confidence: 99%

Generalizing Robot Imitation Learning with Invariant Hidden Semi-Markov Models

Tanwani

Thananjeyan

Laskey

et al. 2019

EasyChair Preprints

Self Cite

View full text Add to dashboard Cite

show abstract

“…However, the parameters {μ, σ } have a big impact on the convergence property. We set the initial values of μ, σ with the K -means method [27], [40].…”

Section: B Initialization Of Task Generative Modelmentioning

confidence: 99%

“…In [25], a parametric hidden Markov model (HMM) was used to encode the data from the demonstrations in the training phase. Tanwani and Calinon [26], [27] developed a task-parameterized HSMM to copy with the environmental situations in the process of manipulation tasks. From the above-mentioned works, the robot learning can be regarded as a problem of feature extraction from the demonstrated training data for a specific skill in the process of HRI.…”

Section: Introductionmentioning

confidence: 99%

Haptics Electromyography Perception and Learning Enhanced Intelligence for Teleoperated Robot

Yang

Liu

et al. 2019

IEEE Trans. Automat. Sci. Eng.

105

View full text Add to dashboard Cite

Due to the lack of transparent and friendly human-robot interaction (HRI) interface, as well as various uncertainties, it is usually a challenge to remotely manipulate a robot to accomplish a complicated task. To improve the teleoperation performance, we propose a new perception mechanism by integrating a novel learning method to operate the robots in the distance. In order to enhance the perception of the teleoperation system, we utilize a surface electromyogram signal to extract the human operator's muscle activation. As a response to the changes in the external environment, as sensed through haptic and visual feedback, a human operator naturally reacts with various muscle activations. By imitating the human behaviors in task execution, not only motion trajectory but also arm stiffness adjusted by muscle activation, it is expected that the robot would be able to carry out the repetitive tasks autonomously or uncertain tasks with improved intelligence. To this end, we develop a robot learning algorithm based on probability statistics under an integrated framework of the hidden semi-Markov model (HSMM) and the Gaussian mixture method. This method is employed to obtain a generative task model based on the robot's trajectory. Then, Gaussian mixture regression based on HSMM is applied to correct the robot trajectory with the reproduced results from the learned task model. The execution procedures consist of a learning phase and a reproduction phase. To guarantee the stability, immersion, and maneuverability of the teleoperation system, a variable gain control method that involves electromyography (EMG) is introduced. Experimental results have demonstrated the effectiveness of the proposed method. Note to Practitioners-This paper is inspired by the limitations of teleoperation to perform a task with unfriendly HRI and lack of intelligence. The human operators need to concentrate

show abstract

“…In [97], a visuo‐haptic AR method was presented to learn object operation tasks, where the HRI system allowed the human operator to choose and operate the virtual objects by adding real workspace. In [98], an online robot learning method with small variance asymptotics was proposed to recognise the human operator's intention and update the telerobot's motion according to the learned task model. For the above‐mentioned approaches, the robot learning methods involve the statistical probability theory and PbD can model the cooperative tasks and copy with the variability the tasks to decrease the workload of humans and work efficiency of the teleoperation system.…”

Section: Key Technologies and Challengesmentioning

confidence: 99%

Combined perception, control, and learning for teleoperation: key technologies, applications, and challenges

Luo

Yang

2020

Cognitive Computation and Systems

View full text Add to dashboard Cite

Teleoperation provides a promising way for human-robot collaboration in the unknown or unstructured environments to perform a cooperative task. It enables humans to complete a task at a remote side and combines both the human's intelligence and the robots' capabilities in a collaborative task. Therefore, it is necessary to conduct cross researches in terms of robotics, artificial intelligence, sensors, and mechatronics. This study covers the state-of-the-arts research in terms of perception, control, and learning. In this review, key technologies about information perception mechanism, control algorithms, human-robot interface, and robot learning strategies for teleoperation are introduced. Then, a comprehensive survey is summarised in representative applications in teleoperation, and the existing challenges and potential directions of the development are discussed.

show abstract

Small-variance asymptotics for non-parametric online robot learning

Cited by 15 publications

References 46 publications

Generalizing Robot Imitation Learning with Invariant Hidden Semi-Markov Models

Generalizing Robot Imitation Learning with Invariant Hidden Semi-Markov Models

Haptics Electromyography Perception and Learning Enhanced Intelligence for Teleoperated Robot

Combined perception, control, and learning for teleoperation: key technologies, applications, and challenges

Contact Info

Product

Resources

About