Probability-Weighted Temporal Registration for Improving Robot Motion Planning and Control Learned from Demonstrations

Bowen, Chris; Alterovitz, Ron

doi:10.1007/978-3-030-44051-0_15

Cited by 1 publication

(1 citation statement)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The interested reader can find connections of probabilistic inference methods for planning and control in Todorov (2008), Toussaint (2009), Kappen et al (2012), and Levine (2018). Other related application contexts in imitation learning include options framework (Fox et al, 2017; Krishnan et al, 2017), sequencing primitives (Medina R. and Billard, 2017), temporal alignment (Bowen and Alterovitz, 2020; Shiarlis et al, 2018), intention recognition (Tanwani and Calinon, 2017; Ti et al, 2019), and neural task programs (Fox et al, 2018; Xu et al, 2017). This paper emphasizes learning manipulation skills from human demonstrations using a family of HMMs by sequencing the atomic movement segments or primitives.…”

Section: Related Workmentioning

confidence: 99%

Sequential robot imitation learning from observations

Tanwani

Yan

Calinon

et al. 2021

The International Journal of Robotics Research

View full text Add to dashboard Cite

This paper presents a framework to learn the sequential structure in the demonstrations for robot imitation learning. We first present a family of task-parameterized hidden semi-Markov models that extracts invariant segments (also called sub-goals or options) from demonstrated trajectories, and optimally follows the sampled sequence of states from the model with a linear quadratic tracking controller. We then extend the concept to learning invariant segments from visual observations that are sequenced together for robot imitation. We present Motion2Vec that learns a deep embedding space by minimizing a metric learning loss in a Siamese network: images from the same action segment are pulled together while being pushed away from randomly sampled images of other segments, and a time contrastive loss is used to preserve the temporal ordering of the images. The trained embeddings are segmented with a recurrent neural network, and subsequently used for decoding the end-effector pose of the robot. We first show its application to a pick-and-place task with the Baxter robot while avoiding a moving obstacle from four kinesthetic demonstrations only, followed by suturing task imitation from publicly available suturing videos of the JIGSAWS dataset with state-of-the-art [Formula: see text]% segmentation accuracy and [Formula: see text] cm error in position per observation on the test set.

show abstract