Probabilistic Consolidation of Grasp Experience
Paper i proceeding, 2016

We present a probabilistic model for joint representation of several sensory modalities and action parameters in a robotic grasping scenario. Our non-linear probabilistic latent variable model encodes relationships between grasp-related parameters, learns the importance of features, and expresses confidence in estimates. The model learns associations between stable and unstable grasps that it experiences during an exploration phase. We demonstrate the applicability of the model for estimating grasp stability, correcting grasps, identifying objects based on tactile imprints and predicting tactile imprints from object-relative gripper poses. We performed experiments on a real platform with both known and novel objects, i.e., objects the robot trained with, and previously unseen objects. Grasp correction had a 75% success rate on known objects, and 73% on new objects. We compared our model to a traditional regression model that succeeded in correcting grasps in only 38% of cases.

Grasping

Stability analysis

Robot sensing systems

Visualization

Grippers

Training

Författare

Yasemin Bekiroglu

University of Birmingham

Andreas Damianou

University of Sheffield

Renaud Detry

Universite de Liège

Kungliga Tekniska Högskolan (KTH)

Johannes Stork

Kungliga Tekniska Högskolan (KTH)

Danica Kragic

Kungliga Tekniska Högskolan (KTH)

Carl Henrik Ek

Kungliga Tekniska Högskolan (KTH)

University of Bristol

Proceedings - IEEE International Conference on Robotics and Automation

10504729 (ISSN)

193-200
978-146738026-3 (ISBN)

2016 IEEE International Conference on Robotics and Automation, ICRA 2016
Stockholm, Sweden,

Ämneskategorier

Robotteknik och automation

Datavetenskap (datalogi)

DOI

10.1109/ICRA.2016.7487133

Mer information

Senast uppdaterat

2022-03-25