Towards Visually-Guided Neuromorphic Robots: Beobots

Towards Visually-Guided Neuromorphic Robots: Beobots. Ng, J., Hirata, R., Mundhenk, T. N., Pichon, E., Tsui, A., Ventrice, T., Williams, P., & Itti, L. In Proc. 9th Joint Symposium on Neural Computation (JSNC'02), Pasadena, California, May, 2002.
abstract bibtex

Despite the advancements made in the field of AI and Robotics, robots today remain vastly inferior to animals in terms of mental agility. The main reason for this is that robots do not possess the neural capabilities of an animal brain. Neural algorithms adapt well to diverse environments, whereas robot AI is usually limited to a test lab setting. To resolve this disparity, an intuitive solution would be to try to emulate the neural functions present in animal brains. However, neural algorithms require vast amounts of computational power to process, in particular those algorithms that require real-time vision. Many robots, which run on power-saving embedded processors, do not have a lot of CPU cycles to spare. We are developing a high-performance visually-guided robotics platform with enough processing speed to run neural algorithms. This 'Beobot' platform consists of a high-performance radio-controlled truck chassis (the robot) carrying an x86-based supercomputer (the Beowulf cluster). The computing cluster consists of two compact dual-CPU motherboards linked together by a gigabit Ethernet connection. Powering the computer are four Pentium-III (Coppermine) 1Ghz processors along with 768MB of memory per motherboard. Two Firewire cameras provide the Beobot's vision. A compact flash card is used as a makeshift hard drive, and it has enough space to store a thin UNIX-variant kernel and iLab's vision software. The vision software itself consists of several general-purpose neural algorithms. Most prominent of these is iLab's Saliency-based visual attention system, which enables the Beobot to drive its attention towards the most salient locations and objects in a visual scene. In addition, we have developed prototype algorithms that allow the Beobot to parse scene layouts and perform object recognition. A primitive action/memory AI system allows it to implement simple visually-guided behavior. Finally, the component-oriented nature of the vision software enables future additions of neural modules. The potential advantage of the Beobot comes from its use of x86-based hardware and UNIX-based C++ development environment. Nearly all the parts of the Beobot are inexpensive, off-the-shelf components. This enables easy replacement of broken parts. Furthermore, the expandability of PC hardware enables devices to be plugged into the Beobot for additional functionalities. All these traits combined make the Beobot potentially easy to replicate, and this allows for wider adoption upon the successful completion of the prototype.

@inproceedings{ Ng_etal02jsnc,
  author = {J. Ng and R. Hirata and T. N. Mundhenk and E. Pichon and A. Tsui
and T. Ventrice and P. Williams and L. Itti},
  title = {Towards Visually-Guided Neuromorphic Robots: Beobots},
  abstract = {Despite the advancements made in the field of AI and
Robotics, robots today remain vastly inferior to animals in terms of
mental agility. The main reason for this is that robots do not possess
the neural capabilities of an animal brain. Neural algorithms adapt
well to diverse environments, whereas robot AI is usually limited to a
test lab setting. To resolve this disparity, an intuitive solution
would be to try to emulate the neural functions present in animal
brains. However, neural algorithms require vast amounts of
computational power to process, in particular those algorithms that
require real-time vision. Many robots, which run on power-saving
embedded processors, do not have a lot of CPU cycles to spare. We are
developing a high-performance visually-guided robotics platform with
enough processing speed to run neural algorithms. This 'Beobot'
platform consists of a high-performance radio-controlled truck chassis
(the robot) carrying an x86-based supercomputer (the Beowulf
cluster). The computing cluster consists of two compact dual-CPU
motherboards linked together by a gigabit Ethernet
connection. Powering the computer are four Pentium-III (Coppermine)
1Ghz processors along with 768MB of memory per motherboard. Two
Firewire cameras provide the Beobot's vision. A compact flash card is
used as a makeshift hard drive, and it has enough space to store a
thin UNIX-variant kernel and iLab's vision software. The vision
software itself consists of several general-purpose neural
algorithms. Most prominent of these is iLab's Saliency-based visual
attention system, which enables the Beobot to drive its attention
towards the most salient locations and objects in a visual scene. In
addition, we have developed prototype algorithms that allow the Beobot
to parse scene layouts and perform object recognition. A primitive
action/memory AI system allows it to implement simple visually-guided
behavior. Finally, the component-oriented nature of the vision
software enables future additions of neural modules. The potential
advantage of the Beobot comes from its use of x86-based hardware and
UNIX-based C++ development environment. Nearly all the parts of the
Beobot are inexpensive, off-the-shelf components. This enables easy
replacement of broken parts. Furthermore, the expandability of PC
hardware enables devices to be plugged into the Beobot for additional
functionalities. All these traits combined make the Beobot potentially
easy to replicate, and this allows for wider adoption upon the
successful completion of the prototype.},
  booktitle = {Proc. 9th Joint Symposium on Neural Computation (JSNC'02),
Pasadena, California},
  year = {2002},
  month = {May},
  type = {mod;bu;cv;bb},
  file = { http://iLab.usc.edu/publications/doc/Ng_etal02jsnc.pdf },
  review = {abs/conf}
}

Downloads: 0

{"_id":{"_str":"5298a1a09eb585cc26000854"},"__v":0,"authorIDs":[],"author_short":["Ng, J.","Hirata, R.","Mundhenk, T.<nbsp>N.","Pichon, E.","Tsui, A.","Ventrice, T.","Williams, P.","Itti, L."],"bibbaseid":"ng-hirata-mundhenk-pichon-tsui-ventrice-williams-itti-towardsvisuallyguidedneuromorphicrobotsbeobots-2002","bibdata":{"html":"<div class=\"bibbase_paper\"> \n\n\n<span class=\"bibbase_paper_titleauthoryear\">\n\t<span class=\"bibbase_paper_title\"><a name=\"Ng_etal02jsnc\"> </a>Towards Visually-Guided Neuromorphic Robots: Beobots.</span>\n\t<span class=\"bibbase_paper_author\">\nNg, J.; Hirata, R.; Mundhenk, T. N.; Pichon, E.; Tsui, A.; Ventrice, T.; Williams, P.; and Itti, L.</span>\n\t\n</span>\n\n\n\nIn\n<i>Proc. 9th Joint Symposium on Neural Computation (JSNC'02), Pasadena, California</i>, May 2002.\n\n\n\n\n\n<br class=\"bibbase_paper_content\"/>\n\n<span class=\"bibbase_paper_content\">\n \n \n \n <a href=\"javascript:showBib('Ng_etal02jsnc')\"\n class=\"bibbase link\">\n \n\t\n\t\n\t\n BibTeX\n <i class=\"fa fa-caret-down\"></i></a>\n \n \n  \n <a class=\"bibbase_abstract_link bibbase link\"\n href=\"javascript:showAbstract('Ng_etal02jsnc')\">\n Abstract\n <i class=\"fa fa-caret-down\"></i></a>\n \n \n \n\n \n \n \n</span>\n\n<div class=\"well well-small bibbase\" id=\"bib_Ng_etal02jsnc\"\n style=\"display:none\">\n <pre>@inproceedings{ Ng_etal02jsnc,\n author = {J. Ng and R. Hirata and T. N. Mundhenk and E. Pichon and A. Tsui\nand T. Ventrice and P. Williams and L. Itti},\n title = {Towards Visually-Guided Neuromorphic Robots: Beobots},\n abstract = {Despite the advancements made in the field of AI and\nRobotics, robots today remain vastly inferior to animals in terms of\nmental agility. The main reason for this is that robots do not possess\nthe neural capabilities of an animal brain. Neural algorithms adapt\nwell to diverse environments, whereas robot AI is usually limited to a\ntest lab setting. To resolve this disparity, an intuitive solution\nwould be to try to emulate the neural functions present in animal\nbrains. However, neural algorithms require vast amounts of\ncomputational power to process, in particular those algorithms that\nrequire real-time vision. Many robots, which run on power-saving\nembedded processors, do not have a lot of CPU cycles to spare. We are\ndeveloping a high-performance visually-guided robotics platform with\nenough processing speed to run neural algorithms. This 'Beobot'\nplatform consists of a high-performance radio-controlled truck chassis\n(the robot) carrying an x86-based supercomputer (the Beowulf\ncluster). The computing cluster consists of two compact dual-CPU\nmotherboards linked together by a gigabit Ethernet\nconnection. Powering the computer are four Pentium-III (Coppermine)\n1Ghz processors along with 768MB of memory per motherboard. Two\nFirewire cameras provide the Beobot's vision. A compact flash card is\nused as a makeshift hard drive, and it has enough space to store a\nthin UNIX-variant kernel and iLab's vision software. The vision\nsoftware itself consists of several general-purpose neural\nalgorithms. Most prominent of these is iLab's Saliency-based visual\nattention system, which enables the Beobot to drive its attention\ntowards the most salient locations and objects in a visual scene. In\naddition, we have developed prototype algorithms that allow the Beobot\nto parse scene layouts and perform object recognition. A primitive\naction/memory AI system allows it to implement simple visually-guided\nbehavior. Finally, the component-oriented nature of the vision\nsoftware enables future additions of neural modules. The potential\nadvantage of the Beobot comes from its use of x86-based hardware and\nUNIX-based C++ development environment. Nearly all the parts of the\nBeobot are inexpensive, off-the-shelf components. This enables easy\nreplacement of broken parts. Furthermore, the expandability of PC\nhardware enables devices to be plugged into the Beobot for additional\nfunctionalities. All these traits combined make the Beobot potentially\neasy to replicate, and this allows for wider adoption upon the\nsuccessful completion of the prototype.},\n booktitle = {Proc. 9th Joint Symposium on Neural Computation (JSNC'02),\nPasadena, California},\n year = {2002},\n month = {May},\n type = {mod;bu;cv;bb},\n file = { http://iLab.usc.edu/publications/doc/Ng_etal02jsnc.pdf },\n review = {abs/conf}\n}</pre>\n</div>\n\n\n<div class=\"well well-small bibbase\" id=\"abstract_Ng_etal02jsnc\"\n style=\"display:none\">\n Despite the advancements made in the field of AI and Robotics, robots today remain vastly inferior to animals in terms of mental agility. The main reason for this is that robots do not possess the neural capabilities of an animal brain. Neural algorithms adapt well to diverse environments, whereas robot AI is usually limited to a test lab setting. To resolve this disparity, an intuitive solution would be to try to emulate the neural functions present in animal brains. However, neural algorithms require vast amounts of computational power to process, in particular those algorithms that require real-time vision. Many robots, which run on power-saving embedded processors, do not have a lot of CPU cycles to spare. We are developing a high-performance visually-guided robotics platform with enough processing speed to run neural algorithms. This 'Beobot' platform consists of a high-performance radio-controlled truck chassis (the robot) carrying an x86-based supercomputer (the Beowulf cluster). The computing cluster consists of two compact dual-CPU motherboards linked together by a gigabit Ethernet connection. Powering the computer are four Pentium-III (Coppermine) 1Ghz processors along with 768MB of memory per motherboard. Two Firewire cameras provide the Beobot's vision. A compact flash card is used as a makeshift hard drive, and it has enough space to store a thin UNIX-variant kernel and iLab's vision software. The vision software itself consists of several general-purpose neural algorithms. Most prominent of these is iLab's Saliency-based visual attention system, which enables the Beobot to drive its attention towards the most salient locations and objects in a visual scene. In addition, we have developed prototype algorithms that allow the Beobot to parse scene layouts and perform object recognition. A primitive action/memory AI system allows it to implement simple visually-guided behavior. Finally, the component-oriented nature of the vision software enables future additions of neural modules. The potential advantage of the Beobot comes from its use of x86-based hardware and UNIX-based C++ development environment. Nearly all the parts of the Beobot are inexpensive, off-the-shelf components. This enables easy replacement of broken parts. Furthermore, the expandability of PC hardware enables devices to be plugged into the Beobot for additional functionalities. All these traits combined make the Beobot potentially easy to replicate, and this allows for wider adoption upon the successful completion of the prototype.\n</div>\n\n\n</div>\n","downloads":0,"bibbaseid":"ng-hirata-mundhenk-pichon-tsui-ventrice-williams-itti-towardsvisuallyguidedneuromorphicrobotsbeobots-2002","role":"author","year":"2002","type":"mod;bu;cv;bb","title":"Towards Visually-Guided Neuromorphic Robots: Beobots","review":"abs/conf","month":"May","key":"Ng_etal02jsnc","id":"Ng_etal02jsnc","file":"http://iLab.usc.edu/publications/doc/Ng_etal02jsnc.pdf","booktitle":"Proc. 9th Joint Symposium on Neural Computation (JSNC'02), Pasadena, California","bibtype":"inproceedings","bibtex":"@inproceedings{ Ng_etal02jsnc,\n author = {J. Ng and R. Hirata and T. N. Mundhenk and E. Pichon and A. Tsui\nand T. Ventrice and P. Williams and L. Itti},\n title = {Towards Visually-Guided Neuromorphic Robots: Beobots},\n abstract = {Despite the advancements made in the field of AI and\nRobotics, robots today remain vastly inferior to animals in terms of\nmental agility. The main reason for this is that robots do not possess\nthe neural capabilities of an animal brain. Neural algorithms adapt\nwell to diverse environments, whereas robot AI is usually limited to a\ntest lab setting. To resolve this disparity, an intuitive solution\nwould be to try to emulate the neural functions present in animal\nbrains. However, neural algorithms require vast amounts of\ncomputational power to process, in particular those algorithms that\nrequire real-time vision. Many robots, which run on power-saving\nembedded processors, do not have a lot of CPU cycles to spare. We are\ndeveloping a high-performance visually-guided robotics platform with\nenough processing speed to run neural algorithms. This 'Beobot'\nplatform consists of a high-performance radio-controlled truck chassis\n(the robot) carrying an x86-based supercomputer (the Beowulf\ncluster). The computing cluster consists of two compact dual-CPU\nmotherboards linked together by a gigabit Ethernet\nconnection. Powering the computer are four Pentium-III (Coppermine)\n1Ghz processors along with 768MB of memory per motherboard. Two\nFirewire cameras provide the Beobot's vision. A compact flash card is\nused as a makeshift hard drive, and it has enough space to store a\nthin UNIX-variant kernel and iLab's vision software. The vision\nsoftware itself consists of several general-purpose neural\nalgorithms. Most prominent of these is iLab's Saliency-based visual\nattention system, which enables the Beobot to drive its attention\ntowards the most salient locations and objects in a visual scene. In\naddition, we have developed prototype algorithms that allow the Beobot\nto parse scene layouts and perform object recognition. A primitive\naction/memory AI system allows it to implement simple visually-guided\nbehavior. Finally, the component-oriented nature of the vision\nsoftware enables future additions of neural modules. The potential\nadvantage of the Beobot comes from its use of x86-based hardware and\nUNIX-based C++ development environment. Nearly all the parts of the\nBeobot are inexpensive, off-the-shelf components. This enables easy\nreplacement of broken parts. Furthermore, the expandability of PC\nhardware enables devices to be plugged into the Beobot for additional\nfunctionalities. All these traits combined make the Beobot potentially\neasy to replicate, and this allows for wider adoption upon the\nsuccessful completion of the prototype.},\n booktitle = {Proc. 9th Joint Symposium on Neural Computation (JSNC'02),\nPasadena, California},\n year = {2002},\n month = {May},\n type = {mod;bu;cv;bb},\n file = { http://iLab.usc.edu/publications/doc/Ng_etal02jsnc.pdf },\n review = {abs/conf}\n}","author_short":["Ng, J.","Hirata, R.","Mundhenk, T.<nbsp>N.","Pichon, E.","Tsui, A.","Ventrice, T.","Williams, P.","Itti, L."],"author":["Ng, J.","Hirata, R.","Mundhenk, T. N.","Pichon, E.","Tsui, A.","Ventrice, T.","Williams, P.","Itti, L."],"abstract":"Despite the advancements made in the field of AI and Robotics, robots today remain vastly inferior to animals in terms of mental agility. The main reason for this is that robots do not possess the neural capabilities of an animal brain. Neural algorithms adapt well to diverse environments, whereas robot AI is usually limited to a test lab setting. To resolve this disparity, an intuitive solution would be to try to emulate the neural functions present in animal brains. However, neural algorithms require vast amounts of computational power to process, in particular those algorithms that require real-time vision. Many robots, which run on power-saving embedded processors, do not have a lot of CPU cycles to spare. We are developing a high-performance visually-guided robotics platform with enough processing speed to run neural algorithms. This 'Beobot' platform consists of a high-performance radio-controlled truck chassis (the robot) carrying an x86-based supercomputer (the Beowulf cluster). The computing cluster consists of two compact dual-CPU motherboards linked together by a gigabit Ethernet connection. Powering the computer are four Pentium-III (Coppermine) 1Ghz processors along with 768MB of memory per motherboard. Two Firewire cameras provide the Beobot's vision. A compact flash card is used as a makeshift hard drive, and it has enough space to store a thin UNIX-variant kernel and iLab's vision software. The vision software itself consists of several general-purpose neural algorithms. Most prominent of these is iLab's Saliency-based visual attention system, which enables the Beobot to drive its attention towards the most salient locations and objects in a visual scene. In addition, we have developed prototype algorithms that allow the Beobot to parse scene layouts and perform object recognition. A primitive action/memory AI system allows it to implement simple visually-guided behavior. Finally, the component-oriented nature of the vision software enables future additions of neural modules. The potential advantage of the Beobot comes from its use of x86-based hardware and UNIX-based C++ development environment. Nearly all the parts of the Beobot are inexpensive, off-the-shelf components. This enables easy replacement of broken parts. Furthermore, the expandability of PC hardware enables devices to be plugged into the Beobot for additional functionalities. All these traits combined make the Beobot potentially easy to replicate, and this allows for wider adoption upon the successful completion of the prototype."},"bibtype":"inproceedings","biburl":"http://ilab.usc.edu/publications/src/ilab.bib","downloads":0,"search_terms":["towards","visually","guided","neuromorphic","robots","beobots","ng","hirata","mundhenk","pichon","tsui","ventrice","williams","itti"],"title":"Towards Visually-Guided Neuromorphic Robots: Beobots","year":2002,"dataSources":["wedBDxEpNXNCLZ2sZ"]}