Papers by Geoffrey Iverson
In this paper, we propose a novel unsupervised color constancy method, called Probabilistic Color... more In this paper, we propose a novel unsupervised color constancy method, called Probabilistic Color Constancy (PCC). We define a framework for estimating the illumination of a scene by weighting the contribution of different image regions using a graph-based representation of the image. To estimate the weight of each (super-)pixel, we rely on two assumptions: (Super-)pixels with similar colors contribute similarly and darker (super-)pixels contribute less. The resulting system has one global optimum solution. The proposed method achieves competitive performance, compared to the state-of-the-art, on INTEL-TAU dataset.
IEEE International Conference on Acoustics Speech and Signal Processing, 1993
Spherical functions arise in geophysics, medical imaging, and computer graphics. This paper addre... more Spherical functions arise in geophysics, medical imaging, and computer graphics. This paper addresses two problems involving spherical functions: determining when two spherical functions are 3-D rotated copies of each other. and averaging several noisy observations of a rotating spherical function. Our solution t,o both problems uses the spherical bispectrum, which is the generalization of the wellknown Euclidean bispectrum. In this paper, we formulate the spherical bispectrum and show that it is invariant under 3-D rotation of the underlying function and unbiased in the presense of additive Gaussian noise. We demonstrate an algorithm for recovering spherical functions from their bispectra.
Journal of the Optical Society of America, Mar 1, 1992
The kth-order autocorrelation function of an image is formed by integrating the product of the im... more The kth-order autocorrelation function of an image is formed by integrating the product of the image and k independently shifted copies of itself: The case k = 1 is the ordinary autocorrelation; k = 2 is the triple correlation. Bartelt et al. [Appl. Opt. 23, 3121 (1984)] have shown that every image of finite size is uniquely determined up to translation by its triple-correlation function. We point out that this is not true in general for images of infinite size, e.g., frequency-band-limited images. Examples are given of pairs of simple bandlimited periodic images and pairs of band-limited aperiodic images that are not translations of each other but that have identical triple correlations. Further examples show that for every k there are distinct band-limited images that have identical kth-order autocorrelation functions. However, certain natural subclasses of infinite images are uniquely determined up to translation by their triple correlations. We develop two general types of criterion for the triple correlation to have an inverse image that is unique up to translation, one based on the zeros of the image spectrum and the other based on image moments. Examples of images satisfying such criteria include diffraction-limited optical images of finite objects and finite images blurred by Gaussian point spreads.
Psychonomic Bulletin & Review, Apr 1, 2010
In their reply, Lecoutre and Killeen (2010) argue for a random effects version of p rep , in whic... more In their reply, Lecoutre and Killeen (2010) argue for a random effects version of p rep , in which the observed effect from one experiment is used to predict the probability that an effect from a different but related experiment will have the same sign. They present a figure giving the impression that this version of p rep accurately predicts the probability of replication. We show that their results are incorrect and conceptually limited, even when corrected. We then present a meaningful evaluation of the random effects p rep as a predictor and find that, as with the fixed effects p rep , it performs very poorly.
Behavioral and Brain Sciences, Jun 1, 1981
Journal of Mathematical Psychology, Dec 1, 1979
ABSTRACT
Journal of the Acoustical Society of America, Apr 1, 1981
The loudness of a tone in a noisy background increases very rapidly above its threshold with incr... more The loudness of a tone in a noisy background increases very rapidly above its threshold with increasing intensity of the tone. The particular curve describing loudness growth as measured by a loudness matching paradigm depends on the level of the background noise. We introduce the hypothesis that an increment in masking noise induces a shift of the loudness matching curve in a diagonal direction (the matching curve being plotted in conventional log-log coordinates). This shift invariance is shown here to hold empiriCally for individual subjects within a monaural loudness matching task, based on a 2IFC paradigm using 1000-Hz tones and a wide band Gaussian noise. Shift invariance places considerable restrictions on models feasible for such data. In conjunction with a very general class of models involving the notion of gain control, the shift invariance property determines all parametric forms possible for the loudness matching functions. The fit of such parametric expressions to the data reported here yields very satisfying results. The resulting model is also found to be consistent with relevant results in the literature.
Journal of Mathematical Psychology, Jun 1, 2006
We re-examine the theoretical status of Fechner's Mathematical Auxiliary Principle [F... more We re-examine the theoretical status of Fechner's Mathematical Auxiliary Principle [Fechner, G. T. (1889). Elemente der psychophysik. Leipzig: Breitkopf und Härtel] which underlies Fechner's method of constructing a sensory scale by integrating just noticeable differences (jnds). That “Principle” has been roundly criticized [Luce, R. D., & Edwards, W. (1958). The derivation of subjective scales from just noticeable differences. Psychological Review,
Mathematical Social Sciences, Oct 1, 1987
ABSTRACT
Abstract : The book is a compilation of the talks given at the 1970 Coral Gables Conference on Fu... more Abstract : The book is a compilation of the talks given at the 1970 Coral Gables Conference on Fundamental Interactions at High Energy, organized by the Center for Theoretical Studies at the University of Miami and held on the University campus. A broadening of outlook has been maintained in these proceedings, which contain current ideas in quantum field theory, S-matrix theory and particle classification and dynamics.
The Journal of the Acoustical Society of America, 1979
Loudness recruitment phenomenon induced by partial masking of a tone by noise was shown to be con... more Loudness recruitment phenomenon induced by partial masking of a tone by noise was shown to be consistent with an automatic gain control type of model [M. Pavel and G. Iverson, J. Acoust. Soc. Am. 63, S18(A) (1978)]. The present study was intended to examine the temporal properties of such an adaptation process. In particular, we were interested in the course of gain recovery following the cessation of a masker. A forward masking paradigm was employed for both the detection of a brief stimulus and for loudness growth determined by loudness matching. The signal was a narrow band click centered at 1 kHz lasting for 10 ms and the masker was a 400 ms rectangular burst of white noise. The 2IFC paradigm was used for both detection and loudness matching. Our analysis focused on two aspects of the data: (1) the relationship of signal level to the masker level—homogeneity (expressed by a “signal to noise” invariance), and (2) the decay of the effect as a function of delay. The homogeneity con...
Journal of Experimental Psychology: Learning, Memory and Cognition, 1991
The mirror effect is a strong regularity in recognition memory: If there are two conditions, A an... more The mirror effect is a strong regularity in recognition memory: If there are two conditions, A and B, with A giving higher recognition accuracy, then old items in A are recognized as old better than old items in B, and also new items in A are recognized as new better than new items in B. The mirror effect is explained by attention/likelihood theory, which also makes several new, counterintuitive predictions. One is that any variable, such as forgetting, that affects recognition changes the responses to new as well as old stimuli. In terms of underlying distributions, forgetting produces concentering, the bilateral movement of distributions, both new (noise) and old (signal), toward a midpoint. Data from two forced-choice experiments are reported that support the prediction of concentering and other predictions drawn from the theory. It is argued that current theories of memory, which are strength theories, cannot handle these regularities.
Psychological Review, 1993
Three regularities in recognition memory are described with supporting data: the mirror effect, t... more Three regularities in recognition memory are described with supporting data: the mirror effect, the order of receiver operating characteristic slopes, and the symmetry of movement of underlying distributions. The derivation of these regularities from attention/likelihood theory is demonstrated. The theory's central concept, which distinguishes it from other theories, is the following: Ss make recognition decisions by combining information about new and old items, the combination made in the form of likelihood ratios. The central role of the likelihood ratios extends the implications of signal detection theory for recognition memory. Attention/likelihood theory is fitted to data of 2 series of experiments. One series involves yes-no tests and confidence ratings, the other forced-choice experiments. It is argued that the regularities require a revision of most current theories of recognition memory.
Optical Society of America Annual Meeting
We discuss the problem of how to determine whether two patterns are identical up to some rotation... more We discuss the problem of how to determine whether two patterns are identical up to some rotation. Much work on this topic in machine vision relies on first-order correlation procedures to determine the magnitudes of the angular-frequency components in the patterns. Because phase information at each frequency is ignored, such strategies identify patterns that are not related by a rotation. How can we capture this phase information, which is needed for fully rotation-invariant recognition? Generalized autocorrelation functions1 provide increasingly more information about relative phases at each angular frequency as the order of the generalized autocorrelation is increased. We examine the conditions on image patterns under which nth-order autocorrelation functions can be used for fully rotation-invariant recognition. The results show that for many applications fully rotation-invariant recognition can be achieved only through generalized correlation procedures that are computationally ...
Advances in Color Vision
Theoretical work on color constancy has dwelt on the problem of estimating surface and illuminant... more Theoretical work on color constancy has dwelt on the problem of estimating surface and illuminant chromatic properties in the case where a set of surfaces is viewed under a single illuminant1-5. Maloney and Wandell4 used linear models to describe variation among surface reflectance functions6-9 and variation among illuminants10,11 to establish a general result for this situation: a trichromatic system viewing surfaces under a single unknown illuminant that is represented by three color descriptors can recover two color- constant descriptors per surface. Their remarkable result is incomplete, however. First, it is desirable to recover three or more descriptors per surface, for neither surface color percepts12 nor surface color properties6-9 are described adequately by two-dimensional models. Second, one would prefer the recovery of surface reflectance descriptors to be independent of the choice of a particular model of illumination. The model that describes accurately the daylight il...
Geometric Representations of Perceptual Phenomena, 2013
Perspectives on Psychological Science, 2011
Statistical inference in psychology has traditionally relied heavily on p-value significance test... more Statistical inference in psychology has traditionally relied heavily on p-value significance testing. This approach to drawing conclusions from data, however, has been widely criticized, and two types of remedies have been advocated. The first proposal is to supplement p values with complementary measures of evidence, such as effect sizes. The second is to replace inference with Bayesian measures of evidence, such as the Bayes factor. The authors provide a practical comparison of p values, effect sizes, and default Bayes factors as measures of statistical evidence, using 855 recently published t tests in psychology. The comparison yields two main results. First, although p values and default Bayes factors almost always agree about what hypothesis is better supported by the data, the measures often disagree about the strength of this support; for 70% of the data sets for which the p value falls between .01 and .05, the default Bayes factor indicates that the evidence is only anecdota...
Uploads
Papers by Geoffrey Iverson