• Corpus ID: 1905737

The student-t mixture as a natural image patch prior with application to image compression

@article{Oord2014TheSM,
  title={The student-t mixture as a natural image patch prior with application to image compression},
  author={A{\"a}ron van den Oord and Benjamin Schrauwen},
  journal={J. Mach. Learn. Res.},
  year={2014},
  volume={15},
  pages={2061-2086}
}
Recent results have shown that Gaussian mixture models (GMMs) are remarkably good at density modeling of natural image patches, especially given their simplicity. In terms of log likelihood on real-valued data they are comparable with the best performing techniques published, easily outperforming more advanced ones, such as deep belief networks. They can be applied to various image processing tasks, such as image denoising, deblurring and inpainting, where they improve on other generic prior… 
View Paper
jmlr.csail.mit.edu
34 Citations
Background Citations
13
Methods Citations
12
Results Citations
1

Figures and Tables from this paper

34 Citations

Citation Type

Citation Type

Factoring Variations in Natural Images with Deep Gaussian Mixture Models
TLDR
This paper proposes a new scalable deep generative model for images, called the Deep Gaussian Mixture Model, that is a straightforward but powerful generalization of GMMs to multiple layers, and shows that deeper GMM architectures generalize better than more shallow ones.
Single Image Super-Resolution via Multiple Mixture Prior Models
TLDR
This paper utilizes an effective way with mixture prior models to transform the large nonlinear feature space of LR images into a group of linear subspaces in the training phase, and finds that student-t prior shows stronger performance than the well-known Gaussian prior.
Image restoration with generalized Gaussian mixture model patch priors
TLDR
It is shown that a generalized Gaussian mixture model (GGMM) captures the underlying distribution of patches better than a GMM and is consistently a better fit for modeling image patch distribution and performs better in image denoising task.
Image Transformer
TLDR
This work generalizes a recently proposed model architecture based on self-attention, the Transformer, to a sequence modeling formulation of image generation with a tractable likelihood, and significantly increases the size of images the model can process in practice, despite maintaining significantly larger receptive fields per layer than typical convolutional neural networks.
Multi-channel expected patch log likelihood for color image denoising
MMPM for Single Image Super-Resolution
The MMPM (Multiple Mixture Prior Model) is learning based SR method which is used for converting Low Resolution (LR) image to High Resolution (HR) image. Other time or space intensive SR methods have…
Conditional Image Generation with PixelCNN Decoders
TLDR
The gated convolutional layers in the proposed model improve the log-likelihood of PixelCNN to match the state-of-the-art performance of PixelRNN on ImageNet, with greatly reduced computational cost.
Generative Image Modeling Using Spatial LSTMs
TLDR
This work introduces a recurrent image model based on multidimensional long short-term memory units which is particularly suited for image modeling due to their spatial structure and outperforms the state of the art in quantitative comparisons on several image datasets and produces promising results when used for texture synthesis and inpainting.
Lossy Image Compression with Compressive Autoencoders
TLDR
It is shown that minimal changes to the loss are sufficient to train deep autoencoders competitive with JPEG 2000 and outperforming recently proposed approaches based on RNNs, and furthermore computationally efficient thanks to a sub-pixel architecture, which makes it suitable for high-resolution images.
Gaussian mixture model learning based image denoising method with adaptive regularization parameters
TLDR
This paper attempts to incorporate the gradient fidelity term with the Gaussian mixture model learning based image denoising method to preserve more fine structures of images and constructs an adaptive regularization parameter selection scheme by combing the image gradient with the local entropy of the image.
...
1
2
3
4
...

References

SHOWING 1-10 OF 40 REFERENCES
Non-local sparse models for image restoration
TLDR
Experimental results in image denoising and demosaicking tasks with synthetic and real noise show that the proposed method outperforms the state of the art, making it possible to effectively restore raw images from digital cameras at a reasonable speed and memory cost.
From learning models of natural image patches to whole image restoration
TLDR
A generic framework which allows for whole image restoration using any patch based prior for which a MAP (or approximate MAP) estimate can be calculated is proposed and a generic, surprisingly simple Gaussian Mixture prior is presented, learned from a set of natural images.
"Natural Images, Gaussian Mixtures and Dead Leaves"
TLDR
An in depth analysis of what such a GMM model learns from natural images as a function of number of mixture components — including covariance structure, contrast variation and intricate structures such as textures, boundaries and more.
Fields of Experts: a framework for learning image priors
  • S. Roth, Michael J. Black
  • Computer Science
    2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05)
  • 2005
TLDR
A framework for learning generic, expressive image priors that capture the statistics of natural scenes and can be used for a variety of machine vision tasks, developed using a Products-of-Experts framework.
Image Denoising Via Sparse and Redundant Representations Over Learned Dictionaries
TLDR
This work addresses the image denoising problem, where zero-mean white and homogeneous Gaussian additive noise is to be removed from a given image, and uses the K-SVD algorithm to obtain a dictionary that describes the image content effectively.
What makes a good model of natural images?
  • Yair Weiss, W. Freeman
  • Computer Science
    2007 IEEE Conference on Computer Vision and Pattern Recognition
  • 2007
TLDR
A tractable lower and upper bounds on the partition function of models based on filter outputs and efficient learning algorithms that do not require any sampling are presented and applied to previous models shows that the nonintuitive features learned are not an artifact of the learning process but rather are capturing robust properties of natural images.
Lloyd clustering of Gauss mixture models for image compression and classification
Mixtures of Conditional Gaussian Scale Mixtures Applied to Multiscale Image Representations
TLDR
It is demonstrated that this probabilistic model based on mixtures of Gaussian scale mixtures and a simple multiscale representation yields the best performance reported to date when evaluated with respect to the cross-entropy rate, a measure tightly linked to the average log-likelihood.
Learning a piecewise linear transform coding scheme for images
TLDR
A piecewise linear transform coding scheme based on Vector Quantization is proposed for the use of GMM's for image compression, demonstrating that the proposed techniques outperform JPEG, with results comparable to JPEG2000 for a broad class of images.
Image Compression Using Sparse Representations and the Iteration-Tuned and Aligned Dictionary
TLDR
It is established experimentally that the ITAD structure results in lower-complexity representations that enjoy greater sparsity when compared to other recent dictionary structures, and a global rate-distortion criterion is proposed that distributes the code bits across the various image blocks.
...
1
2
3
4
...