A Canonical Neural Circuit for Cortical Nonlinear Operations

By Minjoon Kouh, Tomaso A. Poggio. 2008

181 citations recorded by the source corpus.

Subfields: Cognitive Neuroscience, Electrical and Electronic Engineering

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cognitive Neuroscience

In Electrical and Electronic Engineering

Citations in the corpus

How does the brain solve visual object recognition? (2012, neutral): “Notably, both AND-like and OR-like computations can be formulated as variants of the NLNmodel class described above (Kouh and Poggio, 2008), illustrating the link to canonical cortical models (see inset in Figure 6).”
Toward an Integration of Deep Learning and Neuroscience (2016, background): “Neurons with the required Gaussian-like tunings to stored templates could be obtained through canonical, local, normalization-based circuits (Kouh and Poggio, 2008), which can also be tweaked to implement other aspects of a vision architecture like softmax operations and pooling. powerful means of…”
A Normalization Model of Multisensory Integration (2011, unknown): “The apparent ubiquity of divisive normalization in neural circuits 36, 45 makes normalization operating at the level of multisensory integration an attractive general model to account for cross-modal interactions.”
The atoms of neural computation (2014, background): “While we will not attempt a comprehensive survey here, several theories (Anselmi et al., 2013; Bhand, Mudur, Suresh, Saxe, & Ng, 2011; George & Hawkins, 2009; Mallat, 2013; Poggio, Mutch, Leibo, Rosasco, & Tacchetti, 2012) aim to give a general account of a broad range of hierarchical sensory comput”
Neural Implementation of Hierarchical Bayesian Inference by Importance Sampling (2009, background): “It has been suggested that biophysical mechanisms such as shunting inhibition and synaptic depression might account for normalization and gain control [10, 21, 22].”
Value-based gain control: divisive normalization in parietal cortex (2011, background): “…in value coding, the circuit mechanisms generating this representation remain unknown; such computations may emerge from fundamental properties of the neural circuit architecture (Kouh and Poggio, 2008), and identifying the underlying mechanism remains a critical direction for future work.”
Deep Convolutional Networks are Hierarchical Kernel Machines (2015, background): “It is well known that such an operation can be implemented by simple circuits of the lateral inhibition type (see [14]).”
Computing with a Canonical Neural Circuits Model with Pool Normalization and Modulating Feedback (2014, background): “The response gain can be generalized by incorporating a nonlinear transfer function f for the input I, with f (I) = I q , q ∈ N (Kouh & Poggio, 2008).”
Reconstructing Speech from Human Auditory Cortex (2012, background): “Such amplitude-based coding schemes are broadly referred to as ‘‘energy models’’ [32,33].”
The Normalization Model of Attention (2009, background): “Shunting inhibition through lateral connections from other neurons in the cortical neighborhood has been proposed as one possible mechanism for normalization (Carandini and Heeger, 1994; Carandini et al., 1997; Kouh and Poggio, 2008).”
Unsupervised Learning of Invariant Representations in Hierarchical Architectures (2013, background): “for appropriate values of the parameters (see [46]).”