Visual Categorization: How the Monkey Brain Does It
Object categorization is a crucial cognitive ability. It has also received much attention in machine vision. However, the computational processes underlying object categorization in cortex are still poorly understood. In this paper we compare data recorded by Freedman et al. from monkeys to that of view-tuned units in our HMAX model of object recognition in cortex ,. We FInd that the results support a model of object recognition in cortex  in which a population of shape-tuned neurons responding to individual exemplars provides a general basis for neurons tuned to different recognition tasks. Simulations further indicate that this strategy of first learning a general but object class-specific representation as input to a classifier simplifies the learning task. Indeed, the physiological data suggest that in the monkey brain, categorization is performed by PFC neurons performing a simple classification based on the thresholding of a linear sum of the inputs from examplar-tuned units. Such a strategy has various computational advantages, especially with respect to transfer across novel recognition tasks.
KeywordsObject Recognition Receiver Operating Characteristic Model Unit Visual Categorization Monkey Brain
Unable to display preview. Download preview PDF.
- 7.C. Shelton. Three-Dimensional Correspondence. Master’s thesis, MIT, Cambridge, MA, 1996.Google Scholar
- 8.U. Knoblich, D.J. Freedman, and M. Riesenhuber. Categorization in IT and PFC: Model and Experiments. AI Memo 2002-007, CBCL Paper 216, MIT AI Laband CBCL, Cambridge, MA, 2002.Google Scholar
- 9.M. Riesenhuber and T. Poggio. A note on object class representation and categorical perception. AI Memo 1679, CBCL Paper 183, MIT AI Laband CBCL, Cambridge, MA, 1999.Google Scholar
- 10.N.A. Macmillan and C.D. Creelman. Detection Theory: A User’s Guide. 1991.Google Scholar
- 14.S. Edelman. Representation and Recognition in Vision. MIT Press, Cambridge, MA, 1999.Google Scholar