Predictability and redundancy of natural images.

By D. Kersten. 1987

Subfields: Computer Vision and Pattern Recognition

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

Citations in the corpus

Generating Images with Sparse Representations (2021, methodology): “We present a novel Transformer-based autoregressive architecture (Vaswani et al., 2017), called “DCTransformer”, which 1 Kersten (1987) estimated 4-bit grayscale pixel images have (an upper bound of) 1.42 median bits of information per pixel, which means they can generally be compressed to 1 .”
Statistical Modeling and Conceptualization of Visual Patterns (2003, unknown): “Kersten [49], did perhaps, the first experiment measuring the conditional entropy of the intensityatapixelgiventhe intensitiesof itsneighboringpixels, ina spirit similar to Shannon’s [76] experiment of measuring the entropy of English words.”
How Sensitive Is the Human Visual System to the Local Statistics of Natural Images? (2013, background): “Observers can predict extremely local image values better in natural than in random images [27], indicating that the visual system also makes use of local natural image regularities.”
What Is the Goal of Sensory Coding? (1994, background): “…proposed that one must understand the statistical regularity of the mammalian visual environment and its relation to the visual code ( e g , Field 1987, 1989, 1993; Atick and Redlich 1990, 1992; Bialek et al. 1991; Eckert and Buchsbaum 1993; Hancock et al. 1992; van Hateren 1992; Kersten 1987).”
A Parametric Texture Model Based on Joint Statistics of Complex Wavelet Coefficients (2000, background): “The set of all possible images is vast, and yet only a small fraction of these are likely to be encountered in a natural setting (Kersten, 1987; Field, 1987; Daugman, 1989; Ruderman and Bialek, 1994).”
Space and time in visual context (2007, background): “Subjects' knowledge of these correlations is evident in their ability to replace missing pixels in digital image”
Detecting hidden messages using higher-order statistical models (2002, methodology): “In contrast, the approach taken here relies on building higher-order statistical models for natur l images [9, 15, 22, 10, 17] and looking for deviations from these models.”
Detecting Hidden Messages Using Higher-Order Statistics and Support Vector Machines (2002, background): “In contrast, the approach taken here relies on building higher-order statistical models for natural images [13, 19, 29, 14, 21] and looking for deviations from these models.”
On Advances in Statistical Modeling of Natural Images (2004, methodology): “Kersten [47] worked on computing the conditional entropies of the pixel values, given the neighboring pixels in an MRF framework.”
Blind inverse gamma correction (2001, background): “This may prove to be particularly useful for understanding and classifying the statistics of natural images (e.g., [ 14 , 15, 16, 17, 18, 19, 20])”