Towards a simpler model of contour integration in early visual processing using a composite of methods

Towards a simpler model of contour integration in early visual processing using a composite of methods. Mundhenk, T. N. & Itti, L. In Proc. 9th Joint Symposium on Neural Computation (JSNC'02), Pasadena, California, May, 2002.
abstract bibtex

iLab has been attempting to simulate contour integration in early visual preprocessing. Our model starts with a standard butterfly pattern of neural connections that excite or suppress neighboring neurons depending on their preferred visual orientation used for instance by Li (1998). This creates systems where neurons tend to excite other neurons with a collinear orientation, but tend to suppress neurons with a parallel orientation. Our current model attempts to distance itself from many current models that use either neuro synchronization or cascade effect to obtain good contour detection. Instead, we have concentrated on a simpler composite model that uses group suppression gain control, multi scale image analysis and fast plasticity. In this, group suppression works by summing the excitation for small groups of neurons. If the group exceeds threshold, proportionately suppression among the group s neurons is increased. Fast plasticity works by increasing the excitatory ability of a neuron if it has been excited by neighboring neurons to a large enough extent. Finally, multi scale processing works by taking the result of processing the same image in multiple scales on the same neural kernel model at each scale. Experiments on real world images shows that contours are most noticeably improved by the use of group suppression gain control, while tests on computer generated contours provided by Jachen Braun that are of varying size, phase and alignment shows improvement most from the use of fast plasticity and multi scale processing. Our results so far suggest that all three additions a both viable and helpful. Further, our model suggests that simpler mechanisms can be used by the brain in the act of early visual contour integration.

@inproceedings{ Mundhenk_Itti02jsnc,
  author = {T. N. Mundhenk and L. Itti},
  title = {Towards a simpler model of contour integration in early visual
processing using a composite of methods},
  abstract = {iLab has been attempting to simulate contour integration in
early visual preprocessing. Our model starts with a standard butterfly
pattern of neural connections that excite or suppress neighboring
neurons depending on their preferred visual orientation used for
instance by Li (1998). This creates systems where neurons tend to
excite other neurons with a collinear orientation, but tend to
suppress neurons with a parallel orientation. Our current model
attempts to distance itself from many current models that use either
neuro synchronization or cascade effect to obtain good contour
detection. Instead, we have concentrated on a simpler composite model
that uses group suppression gain control, multi scale image analysis
and fast plasticity. In this, group suppression works by summing the
excitation for small groups of neurons. If the group exceeds
threshold, proportionately suppression among the group s neurons is
increased. Fast plasticity works by increasing the excitatory ability
of a neuron if it has been excited by neighboring neurons to a large
enough extent. Finally, multi scale processing works by taking the
result of processing the same image in multiple scales on the same
neural kernel model at each scale. Experiments on real world images
shows that contours are most noticeably improved by the use of group
suppression gain control, while tests on computer generated contours
provided by Jachen Braun that are of varying size, phase and alignment
shows improvement most from the use of fast plasticity and multi scale
processing. Our results so far suggest that all three additions a both
viable and helpful. Further, our model suggests that simpler
mechanisms can be used by the brain in the act of early visual contour
integration. },
  booktitle = {Proc. 9th Joint Symposium on Neural Computation (JSNC'02),
Pasadena, California},
  year = {2002},
  month = {May},
  type = {mod;bu},
  file = { http://iLab.usc.edu/publications/doc/Mundhenk_Itti02jsnc.pdf },
  review = {abs/conf}
}

Downloads: 0

{"_id":{"_str":"5298a1a09eb585cc260008ac"},"__v":0,"authorIDs":[],"author_short":["Mundhenk, T.<nbsp>N.","Itti, L."],"bibbaseid":"mundhenk-itti-towardsasimplermodelofcontourintegrationinearlyvisualprocessingusingacompositeofmethods-2002","bibdata":{"html":"<div class=\"bibbase_paper\"> \n\n\n<span class=\"bibbase_paper_titleauthoryear\">\n\t<span class=\"bibbase_paper_title\"><a name=\"Mundhenk_Itti02jsnc\"> </a>Towards a simpler model of contour integration in early visual processing using a composite of methods.</span>\n\t<span class=\"bibbase_paper_author\">\nMundhenk, T. N.; 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('Mundhenk_Itti02jsnc')\"\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('Mundhenk_Itti02jsnc')\">\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_Mundhenk_Itti02jsnc\"\n style=\"display:none\">\n <pre>@inproceedings{ Mundhenk_Itti02jsnc,\n author = {T. N. Mundhenk and L. Itti},\n title = {Towards a simpler model of contour integration in early visual\nprocessing using a composite of methods},\n abstract = {iLab has been attempting to simulate contour integration in\nearly visual preprocessing. Our model starts with a standard butterfly\npattern of neural connections that excite or suppress neighboring\nneurons depending on their preferred visual orientation used for\ninstance by Li (1998). This creates systems where neurons tend to\nexcite other neurons with a collinear orientation, but tend to\nsuppress neurons with a parallel orientation. Our current model\nattempts to distance itself from many current models that use either\nneuro synchronization or cascade effect to obtain good contour\ndetection. Instead, we have concentrated on a simpler composite model\nthat uses group suppression gain control, multi scale image analysis\nand fast plasticity. In this, group suppression works by summing the\nexcitation for small groups of neurons. If the group exceeds\nthreshold, proportionately suppression among the group s neurons is\nincreased. Fast plasticity works by increasing the excitatory ability\nof a neuron if it has been excited by neighboring neurons to a large\nenough extent. Finally, multi scale processing works by taking the\nresult of processing the same image in multiple scales on the same\nneural kernel model at each scale. Experiments on real world images\nshows that contours are most noticeably improved by the use of group\nsuppression gain control, while tests on computer generated contours\nprovided by Jachen Braun that are of varying size, phase and alignment\nshows improvement most from the use of fast plasticity and multi scale\nprocessing. Our results so far suggest that all three additions a both\nviable and helpful. Further, our model suggests that simpler\nmechanisms can be used by the brain in the act of early visual contour\nintegration. },\n booktitle = {Proc. 9th Joint Symposium on Neural Computation (JSNC'02),\nPasadena, California},\n year = {2002},\n month = {May},\n type = {mod;bu},\n file = { http://iLab.usc.edu/publications/doc/Mundhenk_Itti02jsnc.pdf },\n review = {abs/conf}\n}</pre>\n</div>\n\n\n<div class=\"well well-small bibbase\" id=\"abstract_Mundhenk_Itti02jsnc\"\n style=\"display:none\">\n iLab has been attempting to simulate contour integration in early visual preprocessing. Our model starts with a standard butterfly pattern of neural connections that excite or suppress neighboring neurons depending on their preferred visual orientation used for instance by Li (1998). This creates systems where neurons tend to excite other neurons with a collinear orientation, but tend to suppress neurons with a parallel orientation. Our current model attempts to distance itself from many current models that use either neuro synchronization or cascade effect to obtain good contour detection. Instead, we have concentrated on a simpler composite model that uses group suppression gain control, multi scale image analysis and fast plasticity. In this, group suppression works by summing the excitation for small groups of neurons. If the group exceeds threshold, proportionately suppression among the group s neurons is increased. Fast plasticity works by increasing the excitatory ability of a neuron if it has been excited by neighboring neurons to a large enough extent. Finally, multi scale processing works by taking the result of processing the same image in multiple scales on the same neural kernel model at each scale. Experiments on real world images shows that contours are most noticeably improved by the use of group suppression gain control, while tests on computer generated contours provided by Jachen Braun that are of varying size, phase and alignment shows improvement most from the use of fast plasticity and multi scale processing. Our results so far suggest that all three additions a both viable and helpful. Further, our model suggests that simpler mechanisms can be used by the brain in the act of early visual contour integration.\n</div>\n\n\n</div>\n","downloads":0,"bibbaseid":"mundhenk-itti-towardsasimplermodelofcontourintegrationinearlyvisualprocessingusingacompositeofmethods-2002","role":"author","year":"2002","type":"mod;bu","title":"Towards a simpler model of contour integration in early visual processing using a composite of methods","review":"abs/conf","month":"May","key":"Mundhenk_Itti02jsnc","id":"Mundhenk_Itti02jsnc","file":"http://iLab.usc.edu/publications/doc/Mundhenk_Itti02jsnc.pdf","booktitle":"Proc. 9th Joint Symposium on Neural Computation (JSNC'02), Pasadena, California","bibtype":"inproceedings","bibtex":"@inproceedings{ Mundhenk_Itti02jsnc,\n author = {T. N. Mundhenk and L. Itti},\n title = {Towards a simpler model of contour integration in early visual\nprocessing using a composite of methods},\n abstract = {iLab has been attempting to simulate contour integration in\nearly visual preprocessing. Our model starts with a standard butterfly\npattern of neural connections that excite or suppress neighboring\nneurons depending on their preferred visual orientation used for\ninstance by Li (1998). This creates systems where neurons tend to\nexcite other neurons with a collinear orientation, but tend to\nsuppress neurons with a parallel orientation. Our current model\nattempts to distance itself from many current models that use either\nneuro synchronization or cascade effect to obtain good contour\ndetection. Instead, we have concentrated on a simpler composite model\nthat uses group suppression gain control, multi scale image analysis\nand fast plasticity. In this, group suppression works by summing the\nexcitation for small groups of neurons. If the group exceeds\nthreshold, proportionately suppression among the group s neurons is\nincreased. Fast plasticity works by increasing the excitatory ability\nof a neuron if it has been excited by neighboring neurons to a large\nenough extent. Finally, multi scale processing works by taking the\nresult of processing the same image in multiple scales on the same\nneural kernel model at each scale. Experiments on real world images\nshows that contours are most noticeably improved by the use of group\nsuppression gain control, while tests on computer generated contours\nprovided by Jachen Braun that are of varying size, phase and alignment\nshows improvement most from the use of fast plasticity and multi scale\nprocessing. Our results so far suggest that all three additions a both\nviable and helpful. Further, our model suggests that simpler\nmechanisms can be used by the brain in the act of early visual contour\nintegration. },\n booktitle = {Proc. 9th Joint Symposium on Neural Computation (JSNC'02),\nPasadena, California},\n year = {2002},\n month = {May},\n type = {mod;bu},\n file = { http://iLab.usc.edu/publications/doc/Mundhenk_Itti02jsnc.pdf },\n review = {abs/conf}\n}","author_short":["Mundhenk, T.<nbsp>N.","Itti, L."],"author":["Mundhenk, T. N.","Itti, L."],"abstract":"iLab has been attempting to simulate contour integration in early visual preprocessing. Our model starts with a standard butterfly pattern of neural connections that excite or suppress neighboring neurons depending on their preferred visual orientation used for instance by Li (1998). This creates systems where neurons tend to excite other neurons with a collinear orientation, but tend to suppress neurons with a parallel orientation. Our current model attempts to distance itself from many current models that use either neuro synchronization or cascade effect to obtain good contour detection. Instead, we have concentrated on a simpler composite model that uses group suppression gain control, multi scale image analysis and fast plasticity. In this, group suppression works by summing the excitation for small groups of neurons. If the group exceeds threshold, proportionately suppression among the group s neurons is increased. Fast plasticity works by increasing the excitatory ability of a neuron if it has been excited by neighboring neurons to a large enough extent. Finally, multi scale processing works by taking the result of processing the same image in multiple scales on the same neural kernel model at each scale. Experiments on real world images shows that contours are most noticeably improved by the use of group suppression gain control, while tests on computer generated contours provided by Jachen Braun that are of varying size, phase and alignment shows improvement most from the use of fast plasticity and multi scale processing. Our results so far suggest that all three additions a both viable and helpful. Further, our model suggests that simpler mechanisms can be used by the brain in the act of early visual contour integration."},"bibtype":"inproceedings","biburl":"http://ilab.usc.edu/publications/src/ilab.bib","downloads":0,"search_terms":["towards","simpler","model","contour","integration","early","visual","processing","using","composite","methods","mundhenk","itti"],"title":"Towards a simpler model of contour integration in early visual processing using a composite of methods","year":2002,"dataSources":["wedBDxEpNXNCLZ2sZ"]}