#### DMCA

## What are textons (2002)

Venue: | International Journal of Computer Vision |

Citations: | 82 - 16 self |

### Citations

3532 | A theory for multiresolution signal decomposition: The wavelet representation
- Mallat
- 1989
(Show Context)
Citation Context ... assignment of bases to the classes ϖc etc., and diffusion dynamics which adjust the positions, scales, and orientations of bases and textons. The algorithm is initialized by a matching pursuit method=-=[12]-=- which often yields a very good initial base map B. 2. Replace the integration by importance sampling, and then adjust the parameters Θ through MLE. In general, we can learn the base functions ψ, the ... |

955 | Sparse coding with an overcomplete basis set: A strategy employed by VI
- Olshausen, Field
- 1997
(Show Context)
Citation Context ... recognition. Thirdly, in biological vision the micro-structures in natural images provide an ecological clue for understanding the functions of neurons in the early stage of biological vision systems=-=[1,13]-=-. However, in the literature of computer vision and visual perception, the word “texton” remains a vague concept and a precise mathematical definition has yet to be found. A. Heyden et al. (Eds.): ECC... |

560 | Shiftable multi-scale transforms - Simoncelli, Freeman, et al. - 1992 |

471 |
Possible principles underlying the transformations of sensorymessages.
- Barlow
- 1961
(Show Context)
Citation Context ... recognition. Thirdly, in biological vision the micro-structures in natural images provide an ecological clue for understanding the functions of neurons in the early stage of biological vision systems=-=[1,13]-=-. However, in the literature of computer vision and visual perception, the word “texton” remains a vague concept and a precise mathematical definition has yet to be found. A. Heyden et al. (Eds.): ECC... |

391 |
Textons, the elements of texture perception, and their interactions.
- Julesz
- 1981
(Show Context)
Citation Context ...motion image sequences, which we call movetons. 1 Introduction Texton refers to fundamental micro-structures in generic natural images and the basic elements in early (pre-attentive) visual perception=-=[8]-=-. In practice, the study of textons has important implications in a series of problems. Firstly, decomposing an image into its constituent components reduces information redundancy and, thus, leads to... |

371 |
An information– maximization approach to blind separation and blind deconvolution.
- Bell, Sejnowski
- 1995
(Show Context)
Citation Context ...ng idea is independent component analysis (ICA) which decomposes images as a linear superposition of some image bases which minimizes some measure of dependence between the coefficients of these bases=-=[2]-=-. While the over-complete basis presents a major progress in the pursuit of fundamental image elements, one may wonder what are the image structures beyond bases. By an analogy to physics, if we compa... |

235 | Image compression via joint statistical characterization in the wavelet domain
- Buccigrossi, Simoncelli
- 1999
(Show Context)
Citation Context ...s of natural images. This includes the scale invariance[15], the joint density (histograms) of small image patches (e.g. 3 × 3 pixels)[10,9], and the joint histogram or correlation of filter responses=-=[3]-=-. Then probabilistic models are derived to account for the spatial statistics[7]. The other stream learns over-complete basis from natural images under the general idea of sparse coding[13]. In contra... |

172 | Data compression and harmonic analysis
- Donoho, Vetterli, et al.
- 1998
(Show Context)
Citation Context ...t al. (Eds.): ECCV 2002, LNCS 2353, pp. 793–807, 2002. c○ Springer-Verlag Berlin Heidelberg 2002s794 S.-C. Zhu et al. One related mathematical theory for studying image components is harmonic analysis=-=[5]-=- which is concerned with decomposing some classes of mathematical functions. This includes Fourier transforms, wavelet transforms[4], and recently wedgelets and ridgelet[5] and various image pyramids ... |

167 | Prior learning and Gibbs reactiondiffusion
- Zhu, Mumford
- 1997
(Show Context)
Citation Context ...natural images statistics and image micro-structures, among which two streams are most remarkable. One stream studies the statistical regularities of natural images. This includes the scale invariance=-=[15]-=-, the joint density (histograms) of small image patches (e.g. 3 × 3 pixels)[10,9], and the joint histogram or correlation of filter responses[3]. Then probabilistic models are derived to account for t... |

132 |
Entropy based algorithms for best basis selection
- Coifman, Wickerhauser
- 1992
(Show Context)
Citation Context ...thematical theory for studying image components is harmonic analysis[5] which is concerned with decomposing some classes of mathematical functions. This includes Fourier transforms, wavelet transforms=-=[4]-=-, and recently wedgelets and ridgelet[5] and various image pyramids in image analysis[14]. In recent years, there is a widespread consensus that the optimal set of image components should be learned f... |

130 | Recognizing Surfaces Using Three-Dimensional Textons
- Leung, Malik
- 1999
(Show Context)
Citation Context ...learned from training images as repeating micro-structures. We implement four experiments for comparison. The first experiment computes clusters in feature space of filter responses, as it is done in =-=[11]-=-. This is a discriminative model. The learned cluster centers are transferred into an image icon by pseudo-inverse. Then a typical image structure may appear multiple times as different image icons wi... |

54 |
Transformed component analysis: joint estimation of spatial transformations and image components.
- Frey, Jojic
- 1999
(Show Context)
Citation Context ...on and scaling. To address this problem, we did a second experiment which integrates the clustering withsWhat Are Textons? 795 some affine transform, in an idea of transformed component analysis (TCA)=-=[6]-=-. The number of learned clusters (TCA) is largely reduced, but they lack variability. In the third experiment, we adopt a generative model and assume that an image is generated by a number of bases fr... |

20 | Visual learning by integrating descriptive and generative methods
- GUO, ZHU, et al.
- 2001
(Show Context)
Citation Context ...histograms) of small image patches (e.g. 3 × 3 pixels)[10,9], and the joint histogram or correlation of filter responses[3]. Then probabilistic models are derived to account for the spatial statistics=-=[7]-=-. The other stream learns over-complete basis from natural images under the general idea of sparse coding[13]. In contrast to the orthogonal bases or tight frame in the Fourier and wavelet transforms,... |

15 | Random-Collage Model for Natural Images
- Lee, Huang, et al.
(Show Context)
Citation Context ...re most remarkable. One stream studies the statistical regularities of natural images. This includes the scale invariance[15], the joint density (histograms) of small image patches (e.g. 3 × 3 pixels)=-=[10,9]-=-, and the joint histogram or correlation of filter responses[3]. Then probabilistic models are derived to account for the spatial statistics[7]. The other stream learns over-complete basis from natura... |

4 | Modeling Natural Microimage Statistics
- Koloydenko
- 2000
(Show Context)
Citation Context ...re most remarkable. One stream studies the statistical regularities of natural images. This includes the scale invariance[15], the joint density (histograms) of small image patches (e.g. 3 × 3 pixels)=-=[10,9]-=-, and the joint histogram or correlation of filter responses[3]. Then probabilistic models are derived to account for the spatial statistics[7]. The other stream learns over-complete basis from natura... |