Object recognition via attractors in the human brain


Abstract

One of the most challenging problems faced by the human brain is visual object recognition. We still have a limited understanding of how this comes about, how its invariance is achieved, and how it is implemented in neuronal circuits. Here, we propose a new theory of conscious object recognition in the human brain that is based on attractor dynamics. Each familiar object is considered to have a corresponding attractor stored in the neuronal circuits. The system evolves towards attractors guided by the visual stimulus, and does so even when only partial information about an object is presented (the well-known pattern completion of attractor dynamics). We intend to exploit this property, and using a specially-developed visual recognition paradigm that enables the control of how much visual information is accessible, we intend to investigate whether the recognition-by-attractors theory is valid. The study will be carried out both at psychophysical level and at the brain mechanisms level, with a special emphasis on how gamma oscillations may be involved in attractor dynamics. Methodologically, the present research program will involve: i) development of new visual paradigms suited for the investigation of attractor-based recognition, ii) psychophysical experiments on human subjects, and iii) simultaneous recording of high-density eeg and eye-tracking data. If confirmed, visual perception through attractors is expected to change the way we think about how the brain processes information. First, it would explain why cortical wiring is so highly recurrent, providing circuits ideally suited for attractor dynamics. Second, the existence of attractors would provide direct evidence for how memories, especially those related to visual experience, are stored. Finally, such findings would motivate and guide the search for the recurrent brain loops that implement attractor dynamics and could potentially contribute to a more comprehensive understanding of how the brain works.


Research team

Dr. Eng. Raul C. Muresan, principal investigator. Timeshare on the project > 60%.

[Personal homepage]    [Curriculum Vitae]
Dr. Eng. Vlad V. Moca, postdoc researcher.

[Curriculum Vitae]
Dr. Ioana Tincas, postdoc researcher.

[Curriculum Vitae]


Objectives and activities

  • Activity report for 2010 (in Romanian).
     [pdf]

  • Activity report for 2011 (in Romanian).
     [pdf]

  • Activity report for 2012 (in Romanian).
     [pdf]

  • Activity report for 2013 (in Romanian).
     [pdf]



Results

  • Overview of results for 2010 (in Romanian).
     [pdf]

  • Overview of results for 2011 (in Romanian).
     [pdf]

  • Overview of results for 2012 (in Romanian).
     [pdf]

  • Overview of results for 2013 (in Romanian).
     [pdf]



Publications

Articles:

  • Moca V.V., Nikolic D., Singer W., Muresan R.C. (2013), Membrane resonance enables stable and robust gamma oscillations. Cerebral Cortex (in press).
     [pdf]  [link]

  • Pampu N.C., Vicente R., Muresan R.C., Priesemann V., Siebenhühner F., Wibral M. (2013), Transfer Entropy as a tool for reconstructing interaction delays in neural signals, Proceedings of International Symposium on Signals, Circuits & Systems - ISSCS 2013 (in press).

  • Nikolic D., Muresan R.C., Feng W., Singer W. (2012), Scaled correlation analysis: a better way to compute a cross-correlogram. European Journal of Neuroscience 35(5): 742-762.
     [pdf]  [link]

  • Moca V.V., Tincas I., Melloni L., Muresan R.C. (2011), Visual exploration and object recognition by lattice deformation. PLoS One 6(7): e22831.
     [pdf]  [link]

  • Jurjut O.F., Nikolic D., Singer W., Yu S., Havenith M.N., Muresan R.C. (2011), Timescales of Multineuronal Activity Patterns Reflect Temporal Structure of Visual Stimuli. PLoS One 6(2): e16758.
     [pdf]  [link]

Abstracts, extended abstracts, posters and invited talks:

  • Moca V.V., Muresan R.C. (2013), Discriminating legitimate oscillations from broadband transients, (CNS Meeting 2013) Paris France: BMC Neuroscience 14 (Suppl 1), P286.
     [pdf] [link]

  • Tincas I., Moca V.V., Muresan R.C. (2013), Visual sampling and integration of information in object recognition, Proceedings of the ASSC 17, P2-069, San Diego, July 12-15.

  • Tincas I., Moca V.V., Muresan R.C. (2011), Pupil dilation and visual object recognition, (ICON XI 2011) Palma de Mallorca, Spain: Frontiers in Human Neuroscience, doi:10.3389/conf.fnhum.2011.207.00473.
     [link]

  • Moca V.V., Muresan R.C. (2011), Emergence of beta/gamma oscillations: ING, PING, and what about RING?, (CNS Meeting 2011) Stockholm Sweden: BMC Neuroscience 12 (Suppl 1), p. 230.
     [pdf] [link]

  • Muresan R.C. (2011), Visual Exploration and Object Recognition with the "Dots" Stimuli, invited talk at Castle Ringberg retreat of the Max Planck Institute for Brain Research, Tegernsee, Germany, September 2011.

  • Muresan R.C. (2010), Looking into the brain: where modeling, experiment and analysis meet, invited talk at Diaspora in Cercetarea Stiintifica si Invatamantul Superior din Romania, Workshop Exploratoriu: "Noi perspective de investigare a creierului", Bucharest, Romania, September 2010.