Arm motion reconstruction via feature clustering in joint angle space

Jack DiGiovanna, Justin C. Sanchez, B. J. Fregly, Jose C. Principe

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We hypothesize that a set of movemes can be used to reconstruct bio mechanically realistic movements. Using parameters from a reaching and grasping task we create a representative three-dimensional motion. From this motion we extract features from the joint angle space. We believe that the physiological importance of these features makes them worth investigating as possible movemes. Machine learning techniques are employed to cluster similar features. The clusters are then used to recursively reconstruct the motion trajectory. Even with only twenty clusters, the average trajectory reconstruction error in Cartesian space is less than 1% of the dynamic range of motion. Our ability to create and analyze realistic motions may be crucial to both future BMI experiments where a desired signal is not available and our understanding of motor control.

Original languageEnglish
Title of host publicationIEEE International Conference on Neural Networks - Conference Proceedings
Pages4678-4683
Number of pages6
StatePublished - Dec 1 2006
Externally publishedYes
EventInternational Joint Conference on Neural Networks 2006, IJCNN '06 - Vancouver, BC, Canada
Duration: Jul 16 2006Jul 21 2006

Other

OtherInternational Joint Conference on Neural Networks 2006, IJCNN '06
CountryCanada
CityVancouver, BC
Period7/16/067/21/06

Fingerprint

Trajectories
Learning systems
Experiments

ASJC Scopus subject areas

  • Software

Cite this

DiGiovanna, J., Sanchez, J. C., Fregly, B. J., & Principe, J. C. (2006). Arm motion reconstruction via feature clustering in joint angle space. In IEEE International Conference on Neural Networks - Conference Proceedings (pp. 4678-4683). [1716749]

Arm motion reconstruction via feature clustering in joint angle space. / DiGiovanna, Jack; Sanchez, Justin C.; Fregly, B. J.; Principe, Jose C.

IEEE International Conference on Neural Networks - Conference Proceedings. 2006. p. 4678-4683 1716749.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

DiGiovanna, J, Sanchez, JC, Fregly, BJ & Principe, JC 2006, Arm motion reconstruction via feature clustering in joint angle space. in IEEE International Conference on Neural Networks - Conference Proceedings., 1716749, pp. 4678-4683, International Joint Conference on Neural Networks 2006, IJCNN '06, Vancouver, BC, Canada, 7/16/06.
DiGiovanna J, Sanchez JC, Fregly BJ, Principe JC. Arm motion reconstruction via feature clustering in joint angle space. In IEEE International Conference on Neural Networks - Conference Proceedings. 2006. p. 4678-4683. 1716749
DiGiovanna, Jack ; Sanchez, Justin C. ; Fregly, B. J. ; Principe, Jose C. / Arm motion reconstruction via feature clustering in joint angle space. IEEE International Conference on Neural Networks - Conference Proceedings. 2006. pp. 4678-4683
@inproceedings{903105e17fc847ec89cbb361319eeb3d,
title = "Arm motion reconstruction via feature clustering in joint angle space",
abstract = "We hypothesize that a set of movemes can be used to reconstruct bio mechanically realistic movements. Using parameters from a reaching and grasping task we create a representative three-dimensional motion. From this motion we extract features from the joint angle space. We believe that the physiological importance of these features makes them worth investigating as possible movemes. Machine learning techniques are employed to cluster similar features. The clusters are then used to recursively reconstruct the motion trajectory. Even with only twenty clusters, the average trajectory reconstruction error in Cartesian space is less than 1{\%} of the dynamic range of motion. Our ability to create and analyze realistic motions may be crucial to both future BMI experiments where a desired signal is not available and our understanding of motor control.",
author = "Jack DiGiovanna and Sanchez, {Justin C.} and Fregly, {B. J.} and Principe, {Jose C.}",
year = "2006",
month = "12",
day = "1",
language = "English",
isbn = "0780394909",
pages = "4678--4683",
booktitle = "IEEE International Conference on Neural Networks - Conference Proceedings",

}

TY - GEN

T1 - Arm motion reconstruction via feature clustering in joint angle space

AU - DiGiovanna, Jack

AU - Sanchez, Justin C.

AU - Fregly, B. J.

AU - Principe, Jose C.

PY - 2006/12/1

Y1 - 2006/12/1

N2 - We hypothesize that a set of movemes can be used to reconstruct bio mechanically realistic movements. Using parameters from a reaching and grasping task we create a representative three-dimensional motion. From this motion we extract features from the joint angle space. We believe that the physiological importance of these features makes them worth investigating as possible movemes. Machine learning techniques are employed to cluster similar features. The clusters are then used to recursively reconstruct the motion trajectory. Even with only twenty clusters, the average trajectory reconstruction error in Cartesian space is less than 1% of the dynamic range of motion. Our ability to create and analyze realistic motions may be crucial to both future BMI experiments where a desired signal is not available and our understanding of motor control.

AB - We hypothesize that a set of movemes can be used to reconstruct bio mechanically realistic movements. Using parameters from a reaching and grasping task we create a representative three-dimensional motion. From this motion we extract features from the joint angle space. We believe that the physiological importance of these features makes them worth investigating as possible movemes. Machine learning techniques are employed to cluster similar features. The clusters are then used to recursively reconstruct the motion trajectory. Even with only twenty clusters, the average trajectory reconstruction error in Cartesian space is less than 1% of the dynamic range of motion. Our ability to create and analyze realistic motions may be crucial to both future BMI experiments where a desired signal is not available and our understanding of motor control.

UR - http://www.scopus.com/inward/record.url?scp=40649128319&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=40649128319&partnerID=8YFLogxK

M3 - Conference contribution

SN - 0780394909

SN - 9780780394902

SP - 4678

EP - 4683

BT - IEEE International Conference on Neural Networks - Conference Proceedings

ER -