Modeling image patches with a directed hierarchy of Markov random fields

By Simon Osindero, Geoffrey E. Hinton. 2007

153 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Probabilistic & Bayesian Methods

How researchers see this paper

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

In Computer Vision and Pattern Recognition

In Probabilistic & Bayesian Methods

Citations in the corpus

Deep Generative Image Models using a Laplacian Pyramid of Adversarial Networks (2015, methodology): “Restricted Boltzmann machines [14, 18, 21, 23], Deep Boltzmann machines [26, 8], Denoising auto-encoders [30] all have a generative decoder that reconstructs the image from the latent representation.”
Generative Image Modeling Using Style and Structure Adversarial Networks (2016, background): “Some examples include the deep auto-encoder [22,23] or Restricted Boltzmann machines (RBMs) [24,25,26,27,28].”
Factored 3-Way Restricted Boltzmann Machines For Modeling Natural Images (2010, methodology): “Our model is equivalent to constraining the images in the sequence to be identical. hidden units to control the covariances of the visible units, not just the biases as was done in Osindero and Hinton (2008).”
Learning generative models of mid-level structure in natural images (2012, methodology): “…image structure captured by the models (e.g. Olshausen and Field, 1997; Bell and Sejnowski, 1997; Lewicki and Olshausen, 1999; Hyvärinen and Hoyer, 2000; Roth and Black, 2005; Osindero and Hinton, 2008; Lee et al., 2009; Karklin and Lewicki, 2009; Sinz et al., 2010; Ranzato et al., 2010a,b).”
Modeling pixel means and covariances using factorized third-order boltzmann machines (2010, background): “Osindero and Hinton [23] use a type of RBM in which the binary visible units have direct pairwise interactions that are learned, but these interactions are not modulated by the activities of hidden units: The hidden units only determine the effective biases of the visible units.”
Visualizing Higher-Layer Features of a Deep Network (2009, background): “Often, these filters take the shape of stroke detectors, when trained on digit data, or edge detectors (Gabor filters) when trained on natural image patches (Hinton et al., 2006a; Hinton et al., 2006b; Osindero & Hinton, 2008; Larochelle et al., 2009).”
Exploring Strategies for Training Deep Neural Networks (2009, background): “, 2008), but also in regression (Salakhutdinov and Hinton, 2008), dimensionality reduction (Hinton and Salakhutdinov, 2006; Salakhutdinov and Hinton, 2007b), modeling textures (Osindero and Hinton, 2008), information retrieval (Salakhutdinov and Hinton, 2007a), robotics (Hadsell et al.”
Quantum Boltzmann Machine (2016, methodology): “We emphasize that unlike the classical RBM, in which there are no lateral connections in both hidden and visible layers (for contrastive divergence techniques to work), we only require their absence in the hidden layer, usually called semi-restricted Boltzmann machine [26].”
The Recurrent Temporal Restricted Boltzmann Machine (2008, methodology): “It follows from these equations that the TRBM is a directed graphical model that has an (undirected) RBM at each timeste p (a related directed sequence of Boltzmann Machines has been considered in [7]).”
An Efficient Learning Procedure for Deep Boltzmann Machines (2012, background): “Sample ˜ x t +1 ,i given ˜ x t,i using transition operator T θ t (˜ x t +1 ,i ← ˜ x t,i ) .”
A generative restricted Boltzmann machine based method for high-dimensional motion data modeling (2015, unknown): “To improve the representation power of RBM, semi-restricted Boltzmann machine (SRBM) [11] is introduced to model the lateral interactions between visible variables.”
Understanding Representations Learned in Deep Architectures (2010, background): “Often, these ﬁlters take the shape of stroke detectors, when trained on digit data, or edge detectors (Gabor ﬁlters) when trained on natural image patches (Hinton et al., 2006a,b; Osindero and Hinton, 2008; Larochelle et al., 2009).”