 |
Visual neuron adapt to 1/f statistics |
|||
|
Yuguo YuCarnegie Mellon Links:
Search |
Visual systems have been found to adapt their machinery to process natural signals efficiently [1-6]. What features in the natural stimulus that neurons really care about? One interesting property of natural signal is that its second order statistics is characterized by a 1/f distribution in power spectrum [7-9]. Could this property be important for neuronal encoding? To answer this question, I recorded from V1 neurons of awake monkeys while they are viewing stimuli of different temporal correlations, i.e., 1/f0 , 1/f, and 1/f2, respectively. We found that neurons indeed prefer 1/f signals the best [10]. This provides strong evidence suggesting that this aspect of natural signals is important for neuronal adaptation. The appropriate long-term correlation might be a key driving force in shaping and optimizing the machinery of neurons in their adaptation to the natural environment [10-11]. [1] Simoncelli, E. (2003). Vision and the statistics of the visual environment Curr. Opin. Neurobiol. 13, 144-149. [2] Rieke, F., Bodnar, D.A. and Bialek, W. (1995). Naturalistic stimuli increase the rate and efficiency of information transmission by primary auditory afferents. P.Roy.Soc.Lond.B.Bio. 262, 259-265. [3] Dan, Y., Atick, J.J. and Reid, R.C. (1996). Efficient coding of natural scenes in the lateral geniculate nucleus: experimental test of a computational theory. J.Neurosci. 16, 3351-3362. [4] Baddeley, R., Abbott, L.F., Booth, M.C.A., Sengpiel, F., Freeman, T., Wakeman, E.A. and E. Rolls. (1997). Responses of neurons in primary and inferior temporal visual cortices to natural scenes. P.Roy.Soc.Lond.B.Bio. 264, 1775-1783. [5] Lewen, G.D., Bialek, W. and de Ruyter van Steveninck, R.R. (2001). Neural coding of naturalistic motion stimuli. Network-Comp Neural. 12, 317-329. [6] Vinje, W.E. and Gallant, J.L. (2000). Sparse coding and decorrelation in primary visual cortex during natural vision. Science 287, 1273-1276. [7] Ruderman D.L. and Bialek, W. (1994). Statistics of natural images: scaling in the woods. Phys. Rev. Lett. 73, 814-817. [8] van Hateren, J.H. and van der Schaaf, A. (1996). In SPIE proceedings-Human vision and electronic imaging, Rogowitz, B.E. and Allebach, J.P., ed. (SPIE Press). [9] Soma, R., Nozaki, D., Kwak, S. and Yamamoto, Y. (2004). 1/f Noise Outperforms White Noise in Sensitizing Baroreflex Function in the Human Brain. Phys. Rev. Lett. 91, 078101. (2004). [10] Yu, Y. and Lee, T.S., Adaptation of V1 neurons to 1/f signal statistics, submitted, (2004). [11] Wang, S., Liu, F., Wang, W. and Yu, Y. (2004). Impact of Spatially correlated noise on neuronal firing. Phys. Rev. E 69, 0119091-0119097. 1. Single neuron recording in V1 cortex
2. Recovering the temporal fields (linear and nonlinear kernels) of V1 neurons by Wiener kernel technique:
3. Information theory analysis: Strong's Direct method
|
|||
