AAAI Publications, Second AAAI Conference on Human Computation and Crowdsourcing

Font Size: 
Using Crowdsourcing to Generate Surrogate Training Data for Robotic Grasp Prediction
Matt Unrath, Zhifei Zhang, Alex Goins, Ryan Carpenter, Weng-Keen Wong, Ravi Balasubramanian

Last modified: 2014-09-05

Abstract


As an alternative to the laborious process of collecting training data from physical robotic platforms for learning robotic grasp quality prediction, we explore the use of surrogate training data from crowd-sourced evaluations of images of robotic grasps. We show that in certain regions of the grasp feature space, grasp predictors trained with this surrogate data were almost as accurate as predictors built using data from physical testing with robots.

Keywords


Machine Learning, Crowdsourcing, Robotics

References


Balasubramanian, R.; Xu, L.; Brook, P. D.; Smith, J. R.; and
Matsuoka, Y. 2012. Physical human interactive guidance:
Identifying grasping principles from human-planned grasps.
IEEE Transactions on Robotics 28(4):899–910.

Bentley, J. L. 1975. Multidimensional binary search trees
used for associative searching. Communications of the ACM
18(9):509–517.

Diankov, R., and Kuffner, J. 2008. OpenRAVE: A planning
architecture for autonomous robotics. Technical Report
CMU-RI-TR-08-34, The Robotics Institute, Pittsburgh, PA.

Goins, A. K.; Carpenter, R.; Wong, W.-K.; and Balasubramanian,
R. 2014. Evaluating the efficacy of grasp metrics
for utilization in a gaussian process-based grasp predictor.
In Proceedings of the 2014 IEEE/RSJ International Conference
on Intelligent Robots and Systems.

Jolliffe, I. T. 2002. Principal Component Analysis. New
York: Springer, 2nd edition.

Rasmussen, C. E. 2006. Gaussian processes for machine
learning. MIT Press.


Full Text: PDF