Efficient coding of natural sounds

By Michael S. Lewicki. Nature Neuroscience, 2002

756 citations recorded by the source corpus.

Subfields: Cognitive Neuroscience, Developmental Biology, Signal Processing

View Paper: https://www.semanticscholar.org/paper/4182862a900408d77a071fb4af5a99810937d351

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 Developmental Biology

In Signal Processing

Citations in the corpus

Neural processing of natural sounds (2014, methodology): “One can thus postulate that the physical characteristics of human speech have evolved to be optimally represented at the auditory periphery (while clearly taking into account other constrains) [29].”
Computational mechanisms of sensorimotor control. (2011, neutral): “The third main type of learning, unsupervised learning, has been a focus primarily in the modeling of sensory processing (Lewicki, 2002; Olshausen and Field, 1996).”
Recursive ICA (2006, methodology): “Linear implementations of Barlow’s efficient encoding hypothesis1, such as ICA [1] and sparse coding [2], have been used to explain the very first layers of auditory and visual information processing in the cerebral cortex [1, 2, 3].”
Handbook of Blind Source Separation: Independent Component Analysis and Applications (2010, background): “Sparse coding [41, 70, 69, 1] is closer to the source separation problem and also aims at the learning of a dictionary (redundant or non-redundant) maximizing the sparse representation of a training data set.”
Blind Source Separation and Independent Component Analysis : A Review (2004, neutral): “ICA can be used for finding statistically efficient representations of speech and natural sounds [92, 100].”
Sparse Modeling for Image and Vision Processing (2014, methodology): “the prior term − log p(D); this is the strategy adopted for instance by Lewicki and Olshausen [1999]. Note that optimizing (2.”
Unsupervised learning of auditory filter banks using non-negative matrix factorisation (2008, background): “Lewicki [3] used time-domain ICA without nonnegativity constraints on short segments of natural sounds (including speech) to find the underlying structure.”
Toward a unified theory of efficient, predictive, and sparse coding (2017, background): “Sparse coding has been successful in predicting many aspects of sensory neural responses (23, 24), notably the orientation and motion selectivity of neurons in the primary visual cortex (25–29).”
Sparse coding of sensory inputs. (2004, background): “terms of a strategy to represent natural sounds as sparse independent events [24 ,25].”
Sparse Representation of Sounds in the Unanesthetized Auditory Cortex (2008, background): “For example, it has been suggested that the statistics of natural sensory environments are sparse, and that a sparse code provides a natural match to such environments (Olshausen and Field, 1997; Lewicki, 2002; Olshausen and Field, 2004).”
Divisive Normalization: Justification and Effectiveness as Efficient Coding Transform (2010, background): “Over the years, many works, most notably the ICA methodology, have aimed to find linear efficient coding transforms for natural sensory signals [20, 4, 15].”
Learning a better representation of speech soundwaves using restricted boltzmann machines (2011, background): “such as Independent Components Analysis [1] some progress has been made in learning to extract features directly from the sound wave [2, 3].”
Contrast Gain Control in Auditory Cortex (2011, neutral): “is possible that this preferred distribution is defined by the ensemble of natural sounds (Attneave, 1954; Barlow, 1961; Schwartz and Simoncelli, 2001; Lewicki, 2002).”
Visual perception and the statistical properties of natural scenes. (2008, background): “(For ICA of natural auditory stimuli, see Lewicki 2002.)”
Blind Source Separation and Independent Component Analysis: A Review (2005, methodology): “ICA can be used for finding statistically efficient representations of speech and natural sounds [92, 100].”
Nonlinear Extraction of Independent Components of Natural Images Using Radial Gaussianization (2009, methodology): “ICA methods have shown success in many applications, especially in deriving bases for natural signals (Olshausen & Field, 1996; van der Schaaf & van Hateren, 1996; Bell & Sejnowski, 1997; Lewicki, 2002).”
A Bayesian observer model constrained by efficient coding can explain 'anti-Bayesian' percepts (2015, methodology): “With the efficient coding assumption above, equation (5), we can now express the bias as a function of the prior belief.”