Collaborative Learning of Hierarchical Task Networks from D - PowerPoint Presentation

Slide1

Collaborative Learning of Hierarchical Task Networks from Demonstration and Instruction

Anahita

Mohseni-Kabir

, Sonia

Chernova

and Charles Rich

Worcester Polytechnic InstituteSlide2

Project Objectives and Contributions

Main Goal: Learning complex procedural tasks from human demonstration and

instruction in the form of hierarchical task networks

and applying it to car maintenance domainProject Contributions:Unified system that integrates hierarchical task networks (HTNs) and collaborative discourse theory into the learning from demonstrationLearning task model from a small number of demonstrations Generalization techniquesIntegration of mixed-initiative interaction into the learning process through question asking

2Slide3

Related Work

Collaborative

Discourse

TheoryDisco (ANSI/CEA-2018 standard) (Grosz and Sidner, 1986 and Rich et al., 2001)Learning from DemonstrationMix LfD and planning (Nicolescu and Mataric

, 2003)

Integrate HTN and

LfD (Rybski et al., 2007)Learn from Instruction (Mohan and Laird, 2011)Learn the HTN from task’s traces (Garland et al., 2001)Segmentation (Niekum et al., 2012)Active learning (Cakmak and Thomaz, 2012)

3Slide4

System Architecture

4

Primitive actions

Primitive and Non-primitive actions

Task model visualization

Questions and answersSlide5

Task Structure Learning

Task Hierarchy

Top-Down

Bottom-UpMix of Top-Down and Buttom-UpTemporal ConstraintsSingle demonstrationData flow

5Slide6

System Overview

6Slide7

Generalization

Input Generalization

Part/whole generalization

Type generalizationMerging multiple demonstrations

7

OntologySlide8

System Overview

8Slide9

Question Asking

9

Question Type

Example

Repeated steps

Should I(robot) execute

UnscrewStud

on other objects of type Stud of

LFhub

?

Grouping steps

Should I add a new task with Unscrew and

PutDown

as its steps?Applicability condition of alternative recipesWhat is the applicability condition of Rotate’s recipe with these steps?New task nameWhat is the best name that describes this new task?Input of a taskPlease specify the input of Unscrew.Execution of one of the alternative recipesShould I achieve Rotate by executing recipe1 or recipe2?Slide10

10Slide11

Performance

Tire rotation task

Six primitive actions: Unscrew, Screw, Hang, Unhang,

PutDown and PickUpComplete execution of two recipes of tire rotation requires 128 stepsComplete teaching of the HTN (two recipes) on average requires 26 demonstration interactions

E.g.,

15

demonstrations, 11 instructions, 11 question responses11Slide12

Conclusion and Future Work

Make the interaction as natural as possible by making the UI and robot

look

like a unified systemDo user study and use the real robot instead of the simulationLearn applicability conditions and pre/postconditions of the tasksFailure detection and recovery

12

This work is supported in part by ONR contract N00014-13-1-0735, in collaboration with Dmitry Berenson, Jim

Mainprice , Artem Gritsenko, and Daniel Miller.Slide13

References

Barbara J. Grosz and Candace L.

Sidner

. Attention, intentions, and the structure of discourse. Comput. Linguist., 12(3):175–204, July 1986.Charles Rich, Candace L Sidner, and Neal Lesh. Collagen: applying collaborative discourse theory to human-computer interaction. AI Magazine, 22 (4):15, 2001.Brenna D Argall, Sonia Chernova, Manuela Veloso

, and

Brett Browning. A survey of robot

learning from demonstration. Robotics and Autonomous Systems, 57(5):469–483, 2009.Paul E Rybski, Kevin Yoon, Jeremy Stolarz, and Manuela M Veloso. Interactive robot task training through dialog and demonstration. In ACM/IEEE Int. Conf. on Human-Robot Interaction, pages 49–56

, 2007.

13Slide14

References

Scott

Niekum

, Sarah Osentoski, George Konidaris, and Andrew G Barto. Learning and generalization of complex tasks from unstructured demonstrations. In IEEE/RSJ Int. Conf. on Intelligent Robots and Systems, pages 5239–5246, 2012.Maya Cakmak

and Andrea L

Thomaz

. Designing robot learners that ask good questions. In ACM/IEEE International Conference on Human-Robot Interaction, pages 17–24. ACM, 2012.Monica N Nicolescu and Maja J Mataric. Natural methods for robot task learning: Instructive demonstrations, generalization

and practice. In AAMAS

, pages

241–248, 2003.

14Slide15

Merging

15