Transition state clustering: Unsupervised surgical trajectory segmentation for robot learning

International Journal of Robotics Research - Tập 36 Số 13-14 - Trang 1595-1618 - 2017
Sanjay Krishnan1, Animesh Garg1, Sachin Patil1, Colin Lea2, Gregory D. Hager2, Pieter Abbeel1, Ken Goldberg1
1University of California, Berkeley, USA.
2Johns Hopkins University USA

Tóm tắt

Demonstration trajectories collected from a supervisor in teleoperation are widely used for robot learning, and temporally segmenting the trajectories into shorter, less-variable segments can improve the efficiency and reliability of learning algorithms. Trajectory segmentation algorithms can be sensitive to noise, spurious motions, and temporal variation. We present a new unsupervised segmentation algorithm, transition state clustering (TSC), which leverages repeated demonstrations of a task by clustering segment endpoints across demonstrations. TSC complements any motion-based segmentation algorithm by identifying candidate transitions, clustering them by kinematic similarity, and then correlating the kinematic clusters with available sensory and temporal features. TSC uses a hierarchical Dirichlet process Gaussian mixture model to avoid selecting the number of segments a priori. We present simulated results to suggest that TSC significantly reduces the number of false-positive segments in dynamical systems observed with noise as compared with seven probabilistic and non-probabilistic segmentation algorithms. We additionally compare algorithms that use piecewise linear segment models, and find that TSC recovers segments of a generated piecewise linear trajectory with greater accuracy in the presence of process and observation noise. At the maximum noise level, TSC recovers the ground truth 49% more accurately than alternatives. Furthermore, TSC runs 100× faster than the next most accurate alternative autoregressive models, which require expensive Markov chain Monte Carlo (MCMC)-based inference. We also evaluated TSC on 67 recordings of surgical needle passing and suturing. We supplemented the kinematic recordings with manually annotated visual features that denote grasp and penetration conditions. On this dataset, TSC finds 83% of needle passing transitions and 73% of the suturing transitions annotated by human experts.

Từ khóa


Tài liệu tham khảo

Alvarez M, 2010, Advances in Neural Information Processing Systems, 55

10.1109/TPAMI.2013.86

10.1016/j.robot.2008.10.024

10.1142/S0219843608001431

Barbič J, 2004, Proceedings of Graphics Interface 2004, 185

10.1109/JRA.1986.1087032

10.1109/IROS.2004.1389828

10.1109/MRA.2010.936947

Chiappa S, 2010, Advances in Neural Information Processing Systems, 388

10.1109/COASE.2017.8256258

10.1016/j.robot.2011.07.020

10.1016/0004-3702(72)90051-3

10.1023/A:1013254724861

Fox E, 2009, Advances in Neural Information Processing Systems, 549

Gao Y, 2014, Medical Image Computing and Computer-Assisted Intervention (MICCAI)

10.1109/IROS.2011.6095118

Ghahramani Z, 1993, Advances in Neural Information Processing Systems 6, [7th NIPS Conference], 120

Ijspeert A, 2002, Neural Information Processing Systems (NIPS), 1523

10.1177/0278364911428653

Krishnan S, 2015, International Symposium of Robotics Research

10.1109/MRA.2010.936961

10.1109/ICRA.2012.6224674

10.1109/TRO.2009.2026508

Kulis B, 2012, Proceedings of the 29th International Conference on Machine Learning (ICML 2012)

Lea C, 2015, Proceedings of WACV

10.1007/s10514-014-9397-9

Liang J, Mahler J, Laskey M, Li P, Goldberg K (2017) Using dVRK teleoperation to facilitate deep learning of automation tasks for an industrial robot. Research report, UC Berkeley. Available at: http://goldberg.berkeley.edu/pubs/2017-Liang-DY-Teleop_CASE_CAMERA_READY.pdf

10.1007/11566465_99

10.3109/10929080600989189

10.1109/HUMANOIDS.2014.7041467

10.1109/THMS.2015.2493536

10.1109/HUMANOIDS.2015.7363584

10.1109/HUMANOIDS.2014.7041476

Meier F, 2012, Artificial Intelligence and Statistics, 761

10.1109/IROS.2011.6094676

10.1006/cviu.2000.0897

10.1007/BF00318086

10.1109/ICRA.2015.7139344

Niekum S, 2013, Proceedings of RSS 2013

10.1109/IROS.2012.6386006

Paraschos A, 2013, Advances in Neural Information Processing Systems, 26, 2616

10.1016/j.robot.2012.09.017

10.1109/TMI.2014.2340473

Saeedi A, 2016, arXiv preprint arXiv:1602.06349

Sahbani A, 2008, IROS-2008 Workshop on Grasp and Task Learning By Imitation

10.1007/4-431-31381-8_23

10.1007/s002210050606

10.1109/IROS.2012.6386017

10.1109/LRA.2016.2517825

10.1007/978-3-642-40760-4_43

10.1109/TRO.2010.2065430

10.1109/ICRA.2014.6907291

10.1109/TSMCB.2012.2185694

10.1007/978-3-642-04268-3_53

10.1037/0096-1523.11.6.828

10.1109/ICRA.2015.7139704

10.1109/ICAR.2015.7251510

Willsky AS, 2009, Advances in Neural Information Processing Systems, 549

10.1109/CVPR.2015.7299065

10.1016/j.media.2013.04.007

Thông tin
Thông tin xuất bản

International Journal of Robotics Research

Tập 36 Số 13-14

1595-1618

Thông tin tác giả