von der Malsburg C. 1973. Self-organization of orientation sensitive cells in the striate cortex. Kybernetik 14:85--100.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....of neural development. 53 54 11.2 Future prospects This thesis has explained many basic receptive field properties of neurons in the primary visual cortex. Of course, numerous explanations have been given for V1 receptive fields (see e.g. 10, 28, 94, 121] and cortical topography (e.g. [35, 76, 96, 107, 158], for reviews see [36, 144] in the past. How will we ever be able to say which explanation is the right one In science, models are compared by (a) their explanatory power, and (b) their simplicity. Given two models that equally well describe some given phenomena, the simpler one is usually taken ....

C. von der Malsburg, "Self-organization of orientation-sensitive cells in the striate cortex," Kybernetik, vol. 14, pp. 85--100, 1973.

....of neural development. 53 54 11.2 Future prospects This thesis has explained many basic receptive field properties of neurons in the primary visual cortex. Of course, numerous explanations have been given for V1 receptive fields (see e.g. 10, 28, 94, 121] and cortical topography (e.g. [35, 76, 96, 107, 158], for reviews see [36, 144] in the past. How will we ever be able to say which explanation is the right one In science, models are compared by (a) their explanatory power, and (b) their simplicity. Given two models that equally well describe some given phenomena, the simpler one is usually taken ....

C. von der Malsburg, "Self-organization of orientation-sensitive cells in the striate cortex," Kybernetik, vol. 14, pp. 85--100, 1973.

....or Euclidean distances. Neural versions of the NLM, like Curvilinear Component Analysis (CCA, 3, 4] generally show better performance, particularly when they do not use the traditional Euclidean metrics [9, 13] Finally, nonlinear projection can be achieved by the Self Organizing Map (SOM, [8, 14, 10]) that works with true topology preservation rather than the more constraining distance reproduction. In this framework, Isotop is a new nonlinear projection algorithm combining the advantages of the SOM and the distance preserving algorithms like Sammon s NLM. The following of this paper ....

C. Von Der Malsburg. Self-organization of orientation sensitive cells in the striate cortex. Kybernetik, 14:85--100, 1973. Madrid (Spain), 28-30 August 2002, Springer, Lecture Notes in Computer Science 2415,

.... is a learning rate parameter. Unfortunately, this learning algorithm alone would cause any weight to increase without bound, so some modi cation has to be used to prevent the weights from becoming too large. One possible solution is to limit the absolute values that each weight w i can take [23], while another is to renormalise the weight vector w to have unit length after each update [13] An alternative is to use a weight decay term which causes the weight vector to tend to have unit length as the algorithm progresses, without explicitly normalising it. To see how this works, consider ....

C. von der Malsburg. Self-organization of orientation sensitive cells in the striate cortex. Kybernetik, 14:85-100, 1973.

....in vector notation w = xy: 40) Unfortunately, this learning algorithm alone would cause any weight to increase without bound, so some modi cation has to be used to prevent the weights from becoming too large. One possible solution is to limit the absolute values that each weight w i can take [46], while another is to renormalise the weight vector w to have unit length after each update [23] An alternative is to use a weight decay term which causes the weight vector to tend to have unit length as the algorithm progresses, without explicitly normalising it. To see how this works, consider ....

C. von der Malsburg. Self-organization of orientation sensitive cells in the striate cortex. Kybernetik, 14:85-100, 1973. 33

....of information, and by showing how this procedure may be carried out in a biologically plausible manner. Synaptic changes sub serving learning have traditionally been complemented by neuronally driven normalization processes in the context of self organization of receptive elds and cortical maps [von der Malsburg, 1973, Miller and MacKay, 1994, Goodhill and Barrow, 1994, Sirosh and Miikkulainen, 1994] and continuous unsupervised learning as in principal component analysis networks [Oja, 1982] In these scenarios normalization is necessary to prevent the excessive growth of synaptic ecacies that occurs when ....

....above results show that in order to obtain e ective memory storage, the postsynaptic covariance must be kept negligible. How then may e ective storage take place 7 in the brain with Hebbian learning We now proceed to show that a neuronally driven procedure (essentially similar to that assumed by [von der Malsburg, 1973, Miller and MacKay, 1994] to take place during self organization) can maintain a vanishing covariance and enable e ective memory storage by acting upon ine ective Hebbian synapses and turning them into e ective ones. 3.1 The Neuronal Weight Correction Procedure The solution emerges when ....

[Article contains additional citation context not shown here]

C. von der Malsburg. Self organization of orientation sensitive cells in the striate cortex. Kybernetik, 14:85-100, 1973.

....in a space of MK dimensions where M is the number of synapses for each relaxation neuron. Other possibilities exist for synaptic adaptability. For example, in the classical work of von der Malsburg, synapses are update by normalized Hebbian where the total amount of synaptic e#cacy is fixed [15]. With this approach, the weight space is a manifold of M(K 1) dimensions. The advantage of such loose constraints is that the update can be implemented more locally. We have implemented the normalized Hebbian rule for edge localization. The spatial and orientation attributes are decoded from the ....

C. von der Malsburg. Self-organization of orientation sensitive cells in the striate cortex. Kybernetik, 14:85--100, 1973.

....that constraining or modifying the standard Hebb rule in a particular way will lead to a contrast insensitive tuning width, thus giving an explanation for persistent orientation tuning as observed in the visual cortex. Our analytical results confirm those of simulations done by Von der Malsburg [4] and provide a starting point for further analytical treatment of less restricted stimuli. Key words: orientation tuning, self organising map, Hebbian learning 1 Introduction The Hubel and Wiesel (HW) model for explaining orientation selectivity in the primary visual cortex has been the ....

....of the old rule. Removal of the pattern bias is here found to be equivalent to constraining the dynamics to the hyperplane P i w i ( 0, i.e. to keeping a balance between excitation and inhibition. This is more similar to the constraints used in early computer simulations by Von der Malsburg [4]. The overall majority of the input to the cortex can still be excitatory as we ignored all synapses that are fixed on this time scale. 4 Conclusion Assuming Hebbian plasticity in the LGN V1 connections we have shown analytically that a stable oriention tuned hypercolumn configuration as ....

C. von der Malsburg. Self-organization of orientation sensitive cells in the striate cortex. Kybernetik, 14 (1973) 85--100.

....noisy, features to the most likely context description. Central to our approach is a Kohonen Self Organizing Map (KSOM) 4] that builds up a 2D topological map using the input data. 2. 1 The Kohonen Self Organizing Map The Kohonen Self Organizing Map is based upon earlier work by von der Malsburg [5] and can be classified as a Winner Take All unsupervised learning neural network. It stores prototypes w ij (also known as codebook vectors) of the input vectors at time t, x(t) in the neurons (or cells) of a 2D grid. At each iteration, the neuron that stores the closest prototype to the new ....

von der Malsburg. Self-organization of orientation sensitive cells in the striate cortex. Kybernetic. Vol. 14, Springer- Verlag (1973) 85-100

....cannot provide e ective associative learning in a biologically plausible manner, and must be complemented with neuronally driven synaptic remodeling. The importance of neuronally driven normalization processes has already been demonstrated in the context of self organization of cortical maps [1, 2] and in continuous unsupervised learning as in principal component analysis networks [3] In these scenarios normalization is necessary to prevent the excessive growth of synap tic ecacies that occurs when learning and neuronal activity are strongly coupled. In contradistinction, this paper ....

....above results show that in order to obtain e ective memory storage, the postsynaptic covariance must be kept negligible. How then may e ective storage take place in the brain with Hebbian learning We now proceed to show that a neuronally driven procedure (essentially similar to that assumed by [2, 1] to occur during self organization) can maintain a vanishing covariance and turn ine ective Hebbian synapses into e ective ones. This enables the brain to utilize inecient learning rules which use local information only, but still attain e ective learning capabilities. The solution emerges when ....

[Article contains additional citation context not shown here]

C. von der Malsburg. Self organization of orientation sensitive cells in the striate cortex. Kybernetik, 14:85-100, 1973.

....matching, industrial process monitoring or analysis, faults detection in devices, concept mapping and adaptive routing in telecommunications. 1 Introduction The Kohonen s Self Organizing Maps (SOM) are a kind of artificial neural network historically inspired by sensory maps found in biology ([17, 11, 12]) They are well known for their ability to provide a data mapping where both sample number and dimensionality are reduced, and can be viewed as a non linear extension of Principal Component Analysis (PCA) 2, 15, 14] However, SOM have a major drawback: the mapping is done toward a grid of ....

C. von der Malsburg. Self-organization of orientation sensitive cells in the striate cortex. Kybernetik, 14:85--100, 1973.

....with neuronally driven synaptic remodeling. The importance of neuronally driven normalization processes has already been To whom correspondence should be addressed. Also affiliated with the Sackler school of medicine) demonstrated in the context of self organization of cortical maps [1, 2] and in continuous unsupervised learning as in principal component analysis networks [3] In these scenarios normalization is necessary to prevent the excessive growth of synaptic efficacies that occurs when learning and neuronal activity are strongly coupled. In contradistinction, this paper ....

....above results show that in order to obtain effective memory storage, the postsynaptic covariance must be kept negligible. How then may effective storage take place in the brain with Hebbian learning We now proceed to show that a neuronally driven procedure (essentially similar to that assumed by [2, 1] to occur during self organization) can maintain a vanishing covariance and turn ineffective Hebbian synapses into effective ones. This enables the brain to utilize inefficient learning rules which use local information only, but still attain effective learning capabilities. The solution emerges ....

[Article contains additional citation context not shown here]

C. von der Malsburg. Self organization of orientation sensitive cells in the striate cortex. Kybernetik, 14:85--100, 1973.

No context found.

von der Malsburg C. 1973. Self-organization of orientation sensitive cells in the striate cortex. Kybernetik 14:85--100.

No context found.

C. von der Malsburg. Self-organization of orientation-sensitive cells in the striate cortex. Kybernetik, 15:85--100, 1973.

No context found.

von der Malsburg, Ch. (1973). Self-organization of orientation sensitive cells in the striate cortex. Kybernetik, 14, 85 -- 100.

No context found.

C. von der Malsburg. Self-organization of orientation sensitive cells in the striate cortex. Kybernetik, 14:85100, 1973.

No context found.

C. von der Malsburg, "Self-organization of orientationsensitive cells in the striate cortex," Kybernetik 14,85--100 (1973).

No context found.

von der Malsburg, C. 1973. "Self-organization of orientation sensitive cells in the striate cortex". Kybernetik, 14:85--100.

No context found.

Chr. von der Malsburg. Self-organization of orientation sensitive cells in the striate cortex. Kybernetik, 14:85--100, 1973.

No context found.

C. von der Malsburg. Self-organization of orientation-sensitive cells in the striate cortex. Kybernetik, 15:85--100, 1973.

No context found.

C. von der Malsburg. Self-organization of orientation sensitive cells in the striate cortex. Kybernetik, 14:85--100, 1973.

No context found.

C. von der Malsburg. Self-organization of orientation-sensitive cells in the striate cortex. Kybernetik, 15:85 100, 1973.

No context found.

von der Malsburg, C. (1973) "Self-organization of orientation sensitive cells in the striate cortex" Kybernetik 14, 85-100. 38

No context found.

C. von der Malsburg. Self-organization of orientation sensitive cells in the striate cortex. Kybernetik, 14:85--100, June 1973.

No context found.

Chr. von der Malsburg. Self-organization of orientation sensitive cells in the striate cortex. Kybernetik, 14:85-100, 1973.

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC