Towards Real-World Neurorobotics: Integrated Neuromorphic Visual AttentionCitation formats

  • External authors:
  • Samantha V Adams
  • Alexander D Rast
  • Cameron Patterson
  • Francesco Galluppi
  • Kevin Brohan
  • José-Antonio Pérez-Carrasco
  • Thomas Wennekers
  • Angelo Cangelosi

Standard

Towards Real-World Neurorobotics: Integrated Neuromorphic Visual Attention. / Adams, Samantha V; Rast, Alexander D; Patterson, Cameron; Galluppi, Francesco; Brohan, Kevin; Pérez-Carrasco, José-Antonio; Wennekers, Thomas; Furber, Steve; Cangelosi, Angelo.

21st International Conference, ICONIP 2014, Proceedings, Part III. Vol. 8836 Springer International Publishing Switzerland : Springer Nature, 2014. p. 563-570 (Neural Information Processing, Lecture Notes in Computer Science).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Harvard

Adams, SV, Rast, AD, Patterson, C, Galluppi, F, Brohan, K, Pérez-Carrasco, J-A, Wennekers, T, Furber, S & Cangelosi, A 2014, Towards Real-World Neurorobotics: Integrated Neuromorphic Visual Attention. in 21st International Conference, ICONIP 2014, Proceedings, Part III. vol. 8836, Neural Information Processing, Lecture Notes in Computer Science, Springer Nature, Springer International Publishing Switzerland, pp. 563-570, 21st International Conference, ICONIP 2014, Kuching, Malaysia, 3/11/14. https://doi.org/10.1007/978-3-319-12643-2_68

APA

Adams, S. V., Rast, A. D., Patterson, C., Galluppi, F., Brohan, K., Pérez-Carrasco, J-A., Wennekers, T., Furber, S., & Cangelosi, A. (2014). Towards Real-World Neurorobotics: Integrated Neuromorphic Visual Attention. In 21st International Conference, ICONIP 2014, Proceedings, Part III (Vol. 8836, pp. 563-570). (Neural Information Processing, Lecture Notes in Computer Science). Springer Nature. https://doi.org/10.1007/978-3-319-12643-2_68

Vancouver

Adams SV, Rast AD, Patterson C, Galluppi F, Brohan K, Pérez-Carrasco J-A et al. Towards Real-World Neurorobotics: Integrated Neuromorphic Visual Attention. In 21st International Conference, ICONIP 2014, Proceedings, Part III. Vol. 8836. Springer International Publishing Switzerland: Springer Nature. 2014. p. 563-570. (Neural Information Processing, Lecture Notes in Computer Science). https://doi.org/10.1007/978-3-319-12643-2_68

Author

Adams, Samantha V ; Rast, Alexander D ; Patterson, Cameron ; Galluppi, Francesco ; Brohan, Kevin ; Pérez-Carrasco, José-Antonio ; Wennekers, Thomas ; Furber, Steve ; Cangelosi, Angelo. / Towards Real-World Neurorobotics: Integrated Neuromorphic Visual Attention. 21st International Conference, ICONIP 2014, Proceedings, Part III. Vol. 8836 Springer International Publishing Switzerland : Springer Nature, 2014. pp. 563-570 (Neural Information Processing, Lecture Notes in Computer Science).

Bibtex

@inproceedings{c9ccf8bbd059405093fac2fa1ece24aa,
title = "Towards Real-World Neurorobotics: Integrated Neuromorphic Visual Attention",
abstract = "Neuromorphic hardware and cognitive robots seem like an obvious fit, yet progress to date has been frustrated by a lack of tangible progress in achieving useful real-world behaviour. System limitations: the simple and usually proprietary nature of neuromorphic and robotic platforms, have often been the fundamental barrier. Here we present an integration of a mature “neuromimetic” chip, SpiNNaker, with the humanoid iCub robot using a direct AER - address-event representation - interface that overcomes the need for complex proprietary protocols by sending information as UDP-encoded spikes over an Ethernet link. Using an existing neural model devised for visual object selection, we enable the robot to perform a real-world task: fixating attention upon a selected stimulus. Results demonstrate the effectiveness of interface and model in being able to control the robot towards stimulus-specific object selection. Using SpiNNaker as an embeddable neuromorphic device illustrates the importance of two design features in a prospective neurorobot: universal configurability that allows the chip to be conformed to the requirements of the robot rather than the other way {\textquoteright}round, and standard interfaces that eliminate difficult low-level issues of connectors, cabling, signal voltages, and protocols. While this study is only a building block towards that goal, the iCub-SpiNNaker system demonstrates a path towards meaningful behaviour in robots controlled by neural network chips.",
keywords = "cognitive; robotics; attention; neuromorphic",
author = "Adams, {Samantha V} and Rast, {Alexander D} and Cameron Patterson and Francesco Galluppi and Kevin Brohan and Jos{\'e}-Antonio P{\'e}rez-Carrasco and Thomas Wennekers and Steve Furber and Angelo Cangelosi",
note = "This work was supported under EPSRC Grant EP/J004561/1(BABEL). The SpiNNaker project is supported by EPSRC Grant EP/G015740/1, andby industry partners ARM, Silistix, and Thales.; 21st International Conference, ICONIP 2014 ; Conference date: 03-11-2014 Through 06-11-2014",
year = "2014",
month = nov,
day = "3",
doi = "10.1007/978-3-319-12643-2_68",
language = "English",
isbn = "978-3-319-12642-5",
volume = "8836",
series = "Neural Information Processing, Lecture Notes in Computer Science",
publisher = "Springer Nature",
pages = "563--570",
booktitle = "21st International Conference, ICONIP 2014, Proceedings, Part III",
address = "United States",

}

RIS

TY - GEN

T1 - Towards Real-World Neurorobotics: Integrated Neuromorphic Visual Attention

AU - Adams, Samantha V

AU - Rast, Alexander D

AU - Patterson, Cameron

AU - Galluppi, Francesco

AU - Brohan, Kevin

AU - Pérez-Carrasco, José-Antonio

AU - Wennekers, Thomas

AU - Furber, Steve

AU - Cangelosi, Angelo

N1 - This work was supported under EPSRC Grant EP/J004561/1(BABEL). The SpiNNaker project is supported by EPSRC Grant EP/G015740/1, andby industry partners ARM, Silistix, and Thales.

PY - 2014/11/3

Y1 - 2014/11/3

N2 - Neuromorphic hardware and cognitive robots seem like an obvious fit, yet progress to date has been frustrated by a lack of tangible progress in achieving useful real-world behaviour. System limitations: the simple and usually proprietary nature of neuromorphic and robotic platforms, have often been the fundamental barrier. Here we present an integration of a mature “neuromimetic” chip, SpiNNaker, with the humanoid iCub robot using a direct AER - address-event representation - interface that overcomes the need for complex proprietary protocols by sending information as UDP-encoded spikes over an Ethernet link. Using an existing neural model devised for visual object selection, we enable the robot to perform a real-world task: fixating attention upon a selected stimulus. Results demonstrate the effectiveness of interface and model in being able to control the robot towards stimulus-specific object selection. Using SpiNNaker as an embeddable neuromorphic device illustrates the importance of two design features in a prospective neurorobot: universal configurability that allows the chip to be conformed to the requirements of the robot rather than the other way ’round, and standard interfaces that eliminate difficult low-level issues of connectors, cabling, signal voltages, and protocols. While this study is only a building block towards that goal, the iCub-SpiNNaker system demonstrates a path towards meaningful behaviour in robots controlled by neural network chips.

AB - Neuromorphic hardware and cognitive robots seem like an obvious fit, yet progress to date has been frustrated by a lack of tangible progress in achieving useful real-world behaviour. System limitations: the simple and usually proprietary nature of neuromorphic and robotic platforms, have often been the fundamental barrier. Here we present an integration of a mature “neuromimetic” chip, SpiNNaker, with the humanoid iCub robot using a direct AER - address-event representation - interface that overcomes the need for complex proprietary protocols by sending information as UDP-encoded spikes over an Ethernet link. Using an existing neural model devised for visual object selection, we enable the robot to perform a real-world task: fixating attention upon a selected stimulus. Results demonstrate the effectiveness of interface and model in being able to control the robot towards stimulus-specific object selection. Using SpiNNaker as an embeddable neuromorphic device illustrates the importance of two design features in a prospective neurorobot: universal configurability that allows the chip to be conformed to the requirements of the robot rather than the other way ’round, and standard interfaces that eliminate difficult low-level issues of connectors, cabling, signal voltages, and protocols. While this study is only a building block towards that goal, the iCub-SpiNNaker system demonstrates a path towards meaningful behaviour in robots controlled by neural network chips.

KW - cognitive; robotics; attention; neuromorphic

U2 - 10.1007/978-3-319-12643-2_68

DO - 10.1007/978-3-319-12643-2_68

M3 - Conference contribution

SN - 978-3-319-12642-5

VL - 8836

T3 - Neural Information Processing, Lecture Notes in Computer Science

SP - 563

EP - 570

BT - 21st International Conference, ICONIP 2014, Proceedings, Part III

PB - Springer Nature

CY - Springer International Publishing Switzerland

T2 - 21st International Conference, ICONIP 2014

Y2 - 3 November 2014 through 6 November 2014

ER -