Skip to content

From Pixels to Percepts: Highly Robust Edge Perception and Contour Following Using Deep Learning and an Optical Biomimetic Tactile Sensor

Research output: Contribution to journalArticle

Standard

From Pixels to Percepts : Highly Robust Edge Perception and Contour Following Using Deep Learning and an Optical Biomimetic Tactile Sensor. / Lepora, Nathan F.; Church, Alex; De Kerckhove, Conrad; Hadsell, Raia; Lloyd, John.

In: IEEE Robotics and Automation Letters, Vol. 4, No. 2, 8641397, 01.04.2019, p. 2101-2107.

Research output: Contribution to journalArticle

Harvard

APA

Vancouver

Author

Lepora, Nathan F. ; Church, Alex ; De Kerckhove, Conrad ; Hadsell, Raia ; Lloyd, John. / From Pixels to Percepts : Highly Robust Edge Perception and Contour Following Using Deep Learning and an Optical Biomimetic Tactile Sensor. In: IEEE Robotics and Automation Letters. 2019 ; Vol. 4, No. 2. pp. 2101-2107.

Bibtex

@article{633796a3c1be48fea2825d4eca951fc6,
title = "From Pixels to Percepts: Highly Robust Edge Perception and Contour Following Using Deep Learning and an Optical Biomimetic Tactile Sensor",
abstract = "Deep learning has the potential to have same the impact on robot touch as it has had on robot vision. Optical tactile sensors act as a bridge between the subjects by allowing techniques from vision to be applied to touch. In this letter, we apply deep learning to an optical biomimetic tactile sensor, the TacTip, which images an array of papillae (pins) inside its sensing surface analogous to structures within human skin. Our main result is that the application of a deep convolutional neural network can give reliable edge perception, and, thus a robust policy for planning contact points to move around object contours. Robustness is demonstrated over several irregular and compliant objects with both tapping and continuous sliding, using a model trained only by tapping onto a disk. These results relied on using techniques to encourage generalization to tasks beyond which the model was trained. We expect this is a generic problem in practical applications of tactile sensing that deep learning will solve.",
keywords = "biomimetics, deep learning in robotics and automation, Force and tactile sensing",
author = "Lepora, {Nathan F.} and Alex Church and {De Kerckhove}, Conrad and Raia Hadsell and John Lloyd",
year = "2019",
month = "4",
day = "1",
doi = "10.1109/LRA.2019.2899192",
language = "English",
volume = "4",
pages = "2101--2107",
journal = "IEEE Robotics and Automation Letters",
issn = "2377-3766",
publisher = "Institute of Electrical and Electronics Engineers (IEEE)",
number = "2",

}

RIS - suitable for import to EndNote

TY - JOUR

T1 - From Pixels to Percepts

T2 - Highly Robust Edge Perception and Contour Following Using Deep Learning and an Optical Biomimetic Tactile Sensor

AU - Lepora, Nathan F.

AU - Church, Alex

AU - De Kerckhove, Conrad

AU - Hadsell, Raia

AU - Lloyd, John

PY - 2019/4/1

Y1 - 2019/4/1

N2 - Deep learning has the potential to have same the impact on robot touch as it has had on robot vision. Optical tactile sensors act as a bridge between the subjects by allowing techniques from vision to be applied to touch. In this letter, we apply deep learning to an optical biomimetic tactile sensor, the TacTip, which images an array of papillae (pins) inside its sensing surface analogous to structures within human skin. Our main result is that the application of a deep convolutional neural network can give reliable edge perception, and, thus a robust policy for planning contact points to move around object contours. Robustness is demonstrated over several irregular and compliant objects with both tapping and continuous sliding, using a model trained only by tapping onto a disk. These results relied on using techniques to encourage generalization to tasks beyond which the model was trained. We expect this is a generic problem in practical applications of tactile sensing that deep learning will solve.

AB - Deep learning has the potential to have same the impact on robot touch as it has had on robot vision. Optical tactile sensors act as a bridge between the subjects by allowing techniques from vision to be applied to touch. In this letter, we apply deep learning to an optical biomimetic tactile sensor, the TacTip, which images an array of papillae (pins) inside its sensing surface analogous to structures within human skin. Our main result is that the application of a deep convolutional neural network can give reliable edge perception, and, thus a robust policy for planning contact points to move around object contours. Robustness is demonstrated over several irregular and compliant objects with both tapping and continuous sliding, using a model trained only by tapping onto a disk. These results relied on using techniques to encourage generalization to tasks beyond which the model was trained. We expect this is a generic problem in practical applications of tactile sensing that deep learning will solve.

KW - biomimetics

KW - deep learning in robotics and automation

KW - Force and tactile sensing

UR - http://www.scopus.com/inward/record.url?scp=85062709098&partnerID=8YFLogxK

U2 - 10.1109/LRA.2019.2899192

DO - 10.1109/LRA.2019.2899192

M3 - Article

VL - 4

SP - 2101

EP - 2107

JO - IEEE Robotics and Automation Letters

JF - IEEE Robotics and Automation Letters

SN - 2377-3766

IS - 2

M1 - 8641397

ER -