Interactive Hierarchical Task Learning from a Single Demonstration

Mohseni-Kabir, Anahita; Rich, Charles; Chernova, Sonia; Sidner, Candace L.; Miller, Delbert C.

doi:10.1145/2696454.2696474

Cited by 78 publications

(40 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Human partners who work side-by-side with these cognitive robots are great resources that the robots can directly learn from. Recent years have seen an increasing amount of work on task learning from human partners (Saunders et al, 2006;Chernova and Veloso, 2008;Cantrell et al, 2012;Mohan et al, 2013;Asada et al, 2009;Mohseni-Kabir et al, 2015;Nejati et al, 2006;. Our future work will incorporate interactive learning of verb semantics with task learning to enable autonomy that can learn by communicating with humans.…”

Section: Resultsmentioning

confidence: 99%

Interactive Learning of Grounded Verb Semantics towards Human-Robot Communication

She¹,

Chai²

2017

Proceedings of the 55th Annual Meeting of the Association For Computational Linguistics (Volume 1: Long Papers)

View full text Add to dashboard Cite

To enable human-robot communication and collaboration, previous works represent grounded verb semantics as the potential change of state to the physical world caused by these verbs. Grounded verb semantics are acquired mainly based on the parallel data of the use of a verb phrase and its corresponding sequences of primitive actions demonstrated by humans. The rich interaction between teachers and students that is considered important in learning new skills has not yet been explored. To address this limitation, this paper presents a new interactive learning approach that allows robots to proactively engage in interaction with human partners by asking good questions to learn models for grounded verb semantics. The proposed approach uses reinforcement learning to allow the robot to acquire an optimal policy for its question-asking behaviors by maximizing the long-term reward. Our empirical results have shown that the interactive learning approach leads to more reliable models for grounded verb semantics, especially in the noisy environment which is full of uncertainties. Compared to previous work, the models acquired from interactive learning result in a 48% to 145% performance gain when applied in new situations.

show abstract

Section: Resultsmentioning

confidence: 99%

Interactive Learning of Grounded Verb Semantics towards Human-Robot Communication

She¹,

Chai²

2017

Proceedings of the 55th Annual Meeting of the Association For Computational Linguistics (Volume 1: Long Papers)

View full text Add to dashboard Cite

show abstract

“…The most widely explored approach is Programming by Demonstration (PbD) which involves taking demonstrations of a task as input and inferring the goal of the task or a policy that can be used to accomplish the task [9], [10]. A majority of the work focuses on directly learning a policy or modeling higher level actions from lower level control signals [11], [12], [13], [14], while some explore learning task goals or task structure represented in various ways [15], [16], [17], [18], [19]. Most closely related to our work, Akgun et al explored simultaneous learning of actions and goals by demonstration, focusing on manipulation tasks, such as closing a box and pouring beans into a bowl [20].…”

Section: Related Workmentioning

confidence: 99%

Simultaneous End-User Programming of Goals and Actions for Robotic Shelf Organization

Liang¹,

Pellier²,

Fiorino³

et al. 2018

2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

View full text Add to dashboard Cite

Arrangement of items on shelves in stores or warehouses is a tedious, repetitive task that can be feasible for robots to perform. The diversity of products that are available in stores and the different setups and preferences of each store makes pre-programming a robot for this task extremely challenging. Instead, our work argues for enabling end-users to customize the robot to their specific objects and setup at deployment time by programming it themselves. To that end, this paper contributes (i) a task representation for shelf arrangements based on a large dataset of grocery store shelf images, (ii) a method for inferring goal configurations from user inputs including demonstrations and direct parameter specifications, and (iii) a system implementation of the proposed approach that allows simultaneously learning task goals and actions. We evaluate our goal inference approach with ten different teaching strategies that combine alternative user inputs in different ways on the large dataset of grocery configurations, as well as with real human teachers through an online user study (N=32). We evaluate our full system implemented on a Fetch mobile manipulator on eight benchmark tasks that demonstrate endto-end programming and execution of shelf arrangement tasks.

show abstract

“…Model predictive control approaches have also considered time-varying cost functions and may use sampling to avoid obstacles [20]. A number of methods learn high-level tasks from demonstrations, for example, [2,36]. For execution, these methods may use visual servoing in conjunction with subtask-specific motion planners [18] or more general constraints [31].…”

Section: Related Workmentioning

confidence: 99%

Closed-Loop Global Motion Planning for Reactive, Collision-Free Execution of Learned Tasks

Bowen

Alterovitz

2018

J. Hum.-Robot Interact.

View full text Add to dashboard Cite

We present a robot motion planning approach for performing a learned task while reacting to the movement of obstacles and task-relevant objects. We employ a closed-loop, sampling-based motion planner operating multiple times a second that senses obstacles and task-relevant objects and generates collision-free motion plans based on a learned-task model. The task model is learned from expert demonstrations prior to task execution and is represented as a hidden Markov model. During task execution, our motion planner quickly searches in the Cartesian product of the task model and a probabilistic roadmap for a collision-free plan with features most similar to the demonstrations given the current locations of the task-relevant objects. We accelerate replanning using a fast bidirectional search and by biasing the sampling distribution using information from the learned-task model. We show the efficacy of our approach with the Baxter robot performing two tasks.

show abstract

Interactive Hierarchical Task Learning from a Single Demonstration

Cited by 78 publications

References 15 publications

Interactive Learning of Grounded Verb Semantics towards Human-Robot Communication

Interactive Learning of Grounded Verb Semantics towards Human-Robot Communication

Simultaneous End-User Programming of Goals and Actions for Robotic Shelf Organization

Closed-Loop Global Motion Planning for Reactive, Collision-Free Execution of Learned Tasks

Contact Info

Product

Resources

About