Results 11  20
of
608
An Analysis of SingleLayer Networks in Unsupervised Feature Learning
"... A great deal of research has focused on algorithms for learning features from unlabeled data. Indeed, much progress has been made on benchmark datasets like NORB and CIFAR by employing increasingly complex unsupervised learning algorithms and deep models. In this paper, however, we show that several ..."
Abstract

Cited by 209 (19 self)
 Add to MetaCart
(Show Context)
A great deal of research has focused on algorithms for learning features from unlabeled data. Indeed, much progress has been made on benchmark datasets like NORB and CIFAR by employing increasingly complex unsupervised learning algorithms and deep models. In this paper, however, we show that several very simple factors, such as the number of hidden nodes in the model, may be as important to achieving high performance as the choice of learning algorithm or the depth of the model. Specifically, we will apply several offtheshelf feature learning algorithms (sparse autoencoders, sparse RBMs and Kmeans clustering, Gaussian mixtures) to NORB and CIFAR datasets using only singlelayer networks. We then present a detailed analysis of the effect of changes in the model setup: the receptive field size, number of hidden nodes (features), the stepsize (“stride”) between extracted features, and the effect of whitening. Our results show that large numbers of hidden nodes and dense feature extraction are as critical to achieving high performance as the choice of algorithm itself—so critical, in fact, that when these parameters are pushed to their limits, we are able to achieve stateoftheart performance on both CIFAR and NORB using only a single layer of features. More surprisingly, our best performance is based on Kmeans clustering, which is extremely fast, has no hyperparameters to tune beyond the model structure itself, and is very easy implement. Despite the simplicity of our system, we achieve performance beyond all previously published results on the CIFAR10 and NORB datasets (79.6 % and 97.0 % accuracy respectively). 1
Emergence of Phase and ShiftInvariant Features by Decomposition of Natural Images into Independent Feature Subspaces
, 2000
"... this article, we show that the same principle of independence maximization can explain the emergence of phase and shiftinvariant features, similar to those found in complex cells. This new kind of emergence is obtained by maximizing the independence between norms of projections on linear subspaces ..."
Abstract

Cited by 199 (31 self)
 Add to MetaCart
this article, we show that the same principle of independence maximization can explain the emergence of phase and shiftinvariant features, similar to those found in complex cells. This new kind of emergence is obtained by maximizing the independence between norms of projections on linear subspaces (instead of the independence of simple linear filter outputs). Thenorms of the projections on such "independent feature subspaces" then indicate the values of invariant features
Learning Spatially Localized, PartsBased Representation
, 2001
"... In this paper, we propose a novel method, called local nonnegative matrix factorization (LNMF), for learning spatially localized, partsbased subspace representation of visual patterns. An objective function is defined to impose localization constraint, in addition to the nonnegativity constraint i ..."
Abstract

Cited by 193 (7 self)
 Add to MetaCart
(Show Context)
In this paper, we propose a novel method, called local nonnegative matrix factorization (LNMF), for learning spatially localized, partsbased subspace representation of visual patterns. An objective function is defined to impose localization constraint, in addition to the nonnegativity constraint in the standard NMF [1]. This gives a set of bases which not only allows a nonsubtractive (partbased) representation of images but also manifests localized features. An algorithm is presented for the learning of such basis components. Experimental results are presented to compare LNMF with the NMF and PCA methods for face representation and recognition, which demonstrates advantages of LNMF.
Natural Signal Statistics and Sensory Gain Control
 Nature Neuroscience
, 2001
"... The statistical properties of natural images suggest an optimal form of nonlinear decomposition, in which the image is decomposed using a set of linear filters at a variety of positions, scales and orientations, and these linear responses are then rectified and divided by a weighted sum of rectified ..."
Abstract

Cited by 189 (24 self)
 Add to MetaCart
The statistical properties of natural images suggest an optimal form of nonlinear decomposition, in which the image is decomposed using a set of linear filters at a variety of positions, scales and orientations, and these linear responses are then rectified and divided by a weighted sum of rectified responses of nearby filters. Such divisive normalization models have become widely used in modeling steadystate responses of neurons in primary visual cortex. In addition to providing a surprisingly good characterization of "typical" neurons, the statistically optimal version of the model is consistent with unusual changes in tuning properties of these neurons at different contrast levels. These results suggest that the nonlinear response properties of cortical neurons are not an accident of biophysical implementation, but serve an important functional role.
Incremental Online Learning in High Dimensions
 Neural Computation
, 2005
"... Locally weighted projection regression (LWPR) is a new algorithm for incremental nonlinear function approximation in high dimensional spaces with redundant and irrelevant input dimensions. At its core, it employs nonparametric regression with locally linear models. In order to stay computationally e ..."
Abstract

Cited by 160 (18 self)
 Add to MetaCart
Locally weighted projection regression (LWPR) is a new algorithm for incremental nonlinear function approximation in high dimensional spaces with redundant and irrelevant input dimensions. At its core, it employs nonparametric regression with locally linear models. In order to stay computationally e#cient and numerically robust, each local model performs the regression analysis with a small number of univariate regressions in selected directions in input space in the spirit of partial least squares regression. We discuss when and how local learning techniques can successfully work in high dimensional spaces and review the various techniques for local dimensionality reduction before finally deriving the LWPR algorithm. The properties of LWPR are that it i) learns rapidly with second order learning methods based on incremental training, ii) uses statistically sound stochastic leaveoneout cross validation for learning without the need to memorize training data, iii) adjusts its weighting kernels based only on local information in order to minimize the danger of negative interference of incremental learning, iv) has a computational complexity that is linear in the number of inputs, and v) can deal with a large number of  possibly redundant  inputs, as shown in various empirical evaluations with up to 90 dimensional data sets. For a probabilistic interpretation, predictive variance and confidence intervals are derived. To our knowledge, LWPR is the first truly incremental spatially localized learning method that can successfully and e#ciently operate in very high dimensional spaces.
Representation Learning: A Review and New Perspectives
, 2012
"... The success of machine learning algorithms generally depends on data representation, and we hypothesize that this is because different representations can entangle and hide more or less the different explanatory factors of variation behind the data. Although specific domain knowledge can be used to ..."
Abstract

Cited by 159 (4 self)
 Add to MetaCart
The success of machine learning algorithms generally depends on data representation, and we hypothesize that this is because different representations can entangle and hide more or less the different explanatory factors of variation behind the data. Although specific domain knowledge can be used to help design representations, learning with generic priors can also be used, and the quest for AI is motivating the design of more powerful representationlearning algorithms implementing such priors. This paper reviews recent work in the area of unsupervised feature learning and joint training of deep learning, covering advances in probabilistic models, autoencoders, manifold learning, and deep architectures. This motivates longerterm unanswered questions about the appropriate objectives for learning good representations, for computing representations (i.e., inference), and the geometrical connections between representation learning, density estimation and manifold learning.
Sparse deep belief net model for visual area V2
 Advances in Neural Information Processing Systems 20
, 2008
"... Abstract 1 Motivated in part by the hierarchical organization of the neocortex, a number of recently proposed algorithms have tried to learn hierarchical, or “deep, ” structure from unlabeled data. While several authors have formally or informally compared their algorithms to computations performed ..."
Abstract

Cited by 157 (19 self)
 Add to MetaCart
(Show Context)
Abstract 1 Motivated in part by the hierarchical organization of the neocortex, a number of recently proposed algorithms have tried to learn hierarchical, or “deep, ” structure from unlabeled data. While several authors have formally or informally compared their algorithms to computations performed in visual area V1 (and the cochlea), little attempt has been made thus far to evaluate these algorithms in terms of their fidelity for mimicking computations at deeper levels in the cortical hierarchy. This thesis describes an unsupervised learning model that faithfully mimics certain properties of visual area V2. Specifically, we develop a sparse variant of the deep belief networks described by Hinton et al. (2006). We learn two layers of representation in the network, and demonstrate that the first layer, similar to prior work on sparse coding and ICA, results in localized, oriented, edge filters, similar to the gabor functions known to model simple cell receptive fields in area V1. Further, the second layer in our model encodes various combinations of the first layer responses in the data. Specifically, it picks up both collinear (“contour”) features as well as corners and junctions. More interestingly, in a quantitative comparison, the encoding of these more complex “corner ” features matches well with the results from Ito & Komatsu’s study of neural responses to angular stimuli in area V2 of the macaque. This suggests that our sparse variant of deep belief networks holds promise for modeling more higherorder features that are encoded in visual cortex. Conversely, one may also interpret the results reported here as suggestive that visual area V2 is performing computations on its input similar to those performed in (sparse) deep belief networks. This plausible relationship generates some intriguing hypotheses about V2 computations. 1 This thesis is an extended version of an earlier paper by Honglak Lee, Chaitanya Ekanadham, and Andrew Ng titled “Sparse deep belief net model for visual area V2.” 1
Independent Component Analysis Of Natural Image Sequences Yields Spatiotemporal Filters Similar To Simple Cells In Primary Visual Cortex
 PROC. R. SOC. LOND. B
, 1998
"... ..."
(Show Context)
Probabilistic framework for the adaptation and comparison of image codes
 J. OPT. SOC. AM. A
, 1999
"... We apply a Bayesian method for inferring an optimal basis to the problem of finding efficient image codes for natural scenes. The basis functions learned by the algorithm are oriented and localized in both space and frequency, bearing a resemblance to twodimensional Gabor functions, and increasing ..."
Abstract

Cited by 139 (10 self)
 Add to MetaCart
(Show Context)
We apply a Bayesian method for inferring an optimal basis to the problem of finding efficient image codes for natural scenes. The basis functions learned by the algorithm are oriented and localized in both space and frequency, bearing a resemblance to twodimensional Gabor functions, and increasing the number of basis functions results in a greater sampling density in position, orientation, and scale. These properties also resemble the spatial receptive fields of neurons in the primary visual cortex of mammals, suggesting that the receptivefield structure of these neurons can be accounted for by a general efficient coding principle. The probabilistic framework provides a method for comparing the coding efficiency of different bases objectively by calculating their probability given the observed data or by measuring the entropy of the basis function coefficients. The learned bases are shown to have better coding efficiency than traditional Fourier and wavelet bases. This framework also provides a Bayesian solution to the problems of image denoising and filling in of missing pixels. We demonstrate that the results obtained by applying the learned bases to these problems are improved over those obtained with traditional techniques.
Sun: A Bayesian framework for saliency using natural statistics
 Journal of Vision
, 2008
"... We propose a definition of saliency by considering what the visual system is trying to optimize when directing attention. The resulting model is a Bayesian framework from which bottomup saliency emerges naturally as the selfinformation of visual features, and overall saliency (incorporating topdo ..."
Abstract

Cited by 135 (4 self)
 Add to MetaCart
(Show Context)
We propose a definition of saliency by considering what the visual system is trying to optimize when directing attention. The resulting model is a Bayesian framework from which bottomup saliency emerges naturally as the selfinformation of visual features, and overall saliency (incorporating topdown information with bottomup saliency) emerges as the pointwise mutual information between the features and the target when searching for a target. An implementation of our framework demonstrates that our model’s bottomup saliency maps perform as well as or better than existing algorithms in predicting people’s fixations in free viewing. Unlike existing saliency measures, which depend on the statistics of the particular image being viewed, our measure of saliency is derived from natural image statistics, obtained in advance from a collection of natural images. For this reason, we call our model SUN (Saliency Using Natural statistics). A measure of saliency based on natural image statistics, rather than based on a single test image, provides a straightforward explanation for many search asymmetries observed in humans; the statistics of a single test image lead to predictions that are not consistent with these asymmetries. In our model, saliency is computed locally, which is consistent with the neuroanatomy of the early visual system and results in an efficient algorithm with few free parameters.