Learning curves for Gaussian processes models: fluctuations and universality

Based on a statistical mechanics approach, we develop a method for approximately computing average case learning curves and their sample fluctuations for Gaussian process regression models. We give examples for the Wiener process and show that universal relations (that are independent of the input distribution) between error measures can be derived.

A variational approach to learning curves

We combine the replica approach from statistical physics with a variational approach to analyze learning curves analytically. We apply the method to Gaussian process regression. As a main result we derive approximative relations between empirical error measures, the generalization error and the posterior variance.

Interocular suppression and contrast gain control in human vision

The human visual system combines contrast information from the two eyes to produce a single cyclopean representation of the external world. This task requires both summation of congruent images and inhibition of incongruent images across the eyes. These processes were explored psychophysically using narrowband sinusoidal grating stimuli. Initial experiments focussed on binocular interactions within a single detecting mechanism, using contrast discrimination and contrast matching tasks. Consistent with previous findings, dichoptic presentation produced greater masking than monocular or binocular presentation. Four computational models were compared, two of which performed well on all data sets. Suppression between mechanisms was then investigated, using orthogonal and oblique stimuli. Two distinct suppressive pathways were identified, corresponding to monocular and dichoptic presentation. Both pathways impact prior to binocular summation of signals, and differ in their strengths, tuning, and response to adaptation, consistent with recent single-cell findings in cat. Strikingly, the magnitude of dichoptic masking was found to be spatiotemporally scale invariant, whereas monocular masking was dependent on stimulus speed. Interocular suppression was further explored using a novel manipulation, whereby stimuli were presented in dichoptic antiphase. Consistent with the predictions of a computational model, this produced weaker masking than in-phase presentation. This allowed the bandwidths of suppression to be measured without the complicating factor of additive combination of mask and test. Finally, contrast vision in strabismic amblyopia was investigated. Although amblyopes are generally believed to have impaired binocular vision, binocular summation was shown to be intact when stimuli were normalized for interocular sensitivity differences. An alternative account of amblyopia was developed, in which signals in the affected eye are subject to attenuation and additive noise prior to binocular combination.

Do many hands make light work? How to overcome social loafing and gain motivation in work teams

Purpose – The purpose of this paper is to challenge the assumption that process losses of individuals working in teams are unavoidable. The paper aims to challenge this assumption on the basis of social identity theory and recent research. Design/methodology/approach – The approach adopted in this paper is to review the mainstream literature providing strong evidence for motivation problems of individuals working in groups. Based on more recent literature, innovative ways to overcome these problems are discussed. Findings – A social identity-based analysis and recent findings summarized in this paper show that social loafing can be overcome and that even motivation gains in group work can be expected when groups are important for the individual group members' self-concepts. Practical implications – The paper provides human resource professionals and front-line managers with suggestions as to how individual motivation and performance might be increased when working in teams. Originality/value – The paper contributes to the literature by challenging the existing approach to reducing social loafing, i.e. individualizing workers as much as possible, and proposes a team-based approach instead to overcome motivation problems.

Growth curves of four crustose lichens

The growth curves of four common species of crustose lichens, viz., Buellia aethalea (Ach.) Th. Fr., Lecidea tumida Massai., Rhizocarpon geographicum (L.) DC., and Rhizocarpon reductum Th. Fr. were studied at a site in south Gwynedd, north Wales, UK. Radial growth rates (RGR, mm 1.5 yr-1) were greatest in thalli of R. reductum and least in R. geographicum. Variation in RGR between thalli was greater in B. aethalea and L. tumida than in the species of Rhizocarpon. The relationship between growth rate and thallus diameter was not asymptotic; RGR increasing in smaller thalli to a maximum and then declining in larger diameter thalli. A polynomial curve was fitted to the data; the growth curves being fitted best by a second-order (quadratic) curve, the best fit to this model being shown by B. aethalea. A significant linear regression with a negative slope was also fitted to the growth of the larger thalli of each species. The data suggest that the growth curves of the four crustose lichens differ significantly from the asymptotic curves of foliose lichen species. A phase of declining RGR in larger thalli appears to be characteristic of crustose lichens and is consistent with data from lichenometric studies.

Suppression and summation in contrast gain control for human vision

Over the last ten years our understanding of early spatial vision has improved enormously. The long-standing model of probability summation amongst multiple independent mechanisms with static output nonlinearities responsible for masking is obsolete. It has been replaced by a much more complex network of additive, suppressive, and facilitatory interactions and nonlinearities across eyes, area, spatial frequency, and orientation that extend well beyond the classical recep-tive field (CRF). A review of a substantial body of psychophysical work performed by ourselves (20 papers), and others, leads us to the following tentative account of the processing path for signal contrast. The first suppression stage is monocular, isotropic, non-adaptable, accelerates with RMS contrast, most potent for low spatial and high temporal frequencies, and extends slightly beyond the CRF. Second and third stages of suppression are difficult to disentangle but are possibly pre- and post-binocular summation, and involve components that are scale invariant, isotropic, anisotropic, chromatic, achromatic, adaptable, interocular, substantially larger than the CRF, and saturated by contrast. The monocular excitatory pathways begin with half-wave rectification, followed by a preliminary stage of half-binocular summation, a square-law transducer, full binocular summation, pooling over phase, cross-mechanism facilitatory interactions, additive noise, linear summation over area, and a slightly uncertain decision-maker. The purpose of each of these interactions is far from clear, but the system benefits from area and binocular summation of weak contrast signals as well as area and ocularity invariances above threshold (a herd of zebras doesn't change its contrast when it increases in number or when you close one eye). One of many remaining challenges is to determine the stage or stages of spatial tuning in the excitatory pathway.

Interocular masking and summation indicate two stages of divisive contrast gain control

Our understanding of early spatial vision owes much to contrast masking and summation paradigms. In particular, the deep region of facilitation at low mask contrasts is thought to indicate a rapidly accelerating contrast transducer (eg a square-law or greater). In experiment 1, we tapped an early stage of this process by measuring monocular and binocular thresholds for patches of 1 cycle deg-1 sine-wave grating. Threshold ratios were around 1.7, implying a nearly linear transducer with an exponent around 1.3. With this form of transducer, two previous models (Legge, 1984 Vision Research 24 385 - 394; Meese et al, 2004 Perception 33 Supplement, 41) failed to fit the monocular, binocular, and dichoptic masking functions measured in experiment 2. However, a new model with two-stages of divisive gain control fits the data very well. Stage 1 incorporates nearly linear monocular transducers (to account for the high level of binocular summation and slight dichoptic facilitation), and monocular and interocular suppression (to fit the profound 42 Oral presentations: Spatial vision Thursday dichoptic masking). Stage 2 incorporates steeply accelerating transduction (to fit the deep regions of monocular and binocular facilitation), and binocular summation and suppression (to fit the monocular and binocular masking). With all model parameters fixed from the discrimination thresholds, we examined the slopes of the psychometric functions. The monocular and binocular slopes were steep (Weibull ߘ3-4) at very low mask contrasts and shallow (ߘ1.2) at all higher contrasts, as predicted by all three models. The dichoptic slopes were steep (ߘ3-4) at very low contrasts, and very steep (ß>5.5) at high contrasts (confirming Meese et al, loco cit.). A crucial new result was that intermediate dichoptic mask contrasts produced shallow slopes (ߘ2). Only the two-stage model predicted the observed pattern of slope variation, so providing good empirical support for a two-stage process of binocular contrast transduction. [Supported by EPSRC GR/S74515/01]

Contrast discrimination and pattern masking: contrast gain control with fixed additive noise

We studied the visual mechanisms that serve to encode spatial contrast at threshold and supra-threshold levels. In a 2AFC contrast-discrimination task, observers had to detect the presence of a vertical 1 cycle deg-1 test grating (of contrast dc) that was superimposed on a similar vertical 1 cycle deg-1 pedestal grating, whereas in pattern masking the test grating was accompanied by a very different masking grating (horizontal 1 cycle deg-1, or oblique 3 cycles deg-1). When expressed as threshold contrast (dc at 75% correct) versus mask contrast (c) our results confirm previous ones in showing a characteristic 'dipper function' for contrast discrimination but a smoothly increasing threshold for pattern masking. However, fresh insight is gained by analysing and modelling performance (p; percent correct) as a joint function of (c, dc) - the performance surface. In contrast discrimination, psychometric functions (p versus logdc) are markedly less steep when c is above threshold, but in pattern masking this reduction of slope did not occur. We explored a standard gain-control model with six free parameters. Three parameters control the contrast response of the detection mechanism and one parameter weights the mask contrast in the cross-channel suppression effect. We assume that signal-detection performance (d') is limited by additive noise of constant variance. Noise level and lapse rate are also fitted parameters of the model. We show that this model accounts very accurately for the whole performance surface in both types of masking, and thus explains the threshold functions and the pattern of variation in psychometric slopes. The cross-channel weight is about 0.20. The model shows that the mechanism response to contrast increment (dc) is linearised by the presence of pedestal contrasts but remains nonlinear in pattern masking.

Motion sharpening and contrast:Gain control precedes compressive non-linearity?

Blurred edges appear sharper in motion than when they are stationary. We (Vision Research 38 (1998) 2108) have previously shown how such distortions in perceived edge blur may be accounted for by a model which assumes that luminance contrast is encoded by a local contrast transducer whose response becomes progressively more compressive as speed increases. If the form of the transducer is fixed (independent of contrast) for a given speed, then a strong prediction of the model is that motion sharpening should increase with increasing contrast. We measured the sharpening of periodic patterns over a large range of contrasts, blur widths and speeds. The results indicate that whilst sharpening increases with speed it is practically invariant with contrast. The contrast invariance of motion sharpening is not explained by an early, static compressive non-linearity alone. However, several alternative explanations are also inconsistent with these results. We show that if a dynamic contrast gain control precedes the static non-linear transducer then motion sharpening, its speed dependence, and its invariance with contrast, can be predicted with reasonable accuracy. © 2003 Elsevier Science Ltd. All rights reserved.

A model for motion sharpening: contrast gain control precedes compressive nonlinearity

Blurred edges appear sharper in motion than when they are stationary. We have previously shown how such distortions in perceived edge blur may be explained by a model which assumes that luminance contrast is encoded by a local contrast transducer whose response becomes progressively more compressive as speed increases. To test this model further, we measured the sharpening of drifting, periodic patterns over a large range of contrasts, blur widths, and speeds Human Vision. The results indicate that, while sharpening increased with speed, it was practically invariant with contrast. This contrast invariance cannot be explained by a fixed compressive nonlinearity since that predicts almost no sharpening at low contrasts.We show by computational modelling of spatiotemporal responses that, if a dynamic contrast gain control precedes the static nonlinear transducer, then motion sharpening, its speed dependence, and its invariance with contrast can be predicted with reasonable accuracy.

Optimizing measurement of subjective amplitude of accommodation with defocus curves

PURPOSE: To determine whether letter sequences and/or lens-presentation order should be randomized when measuring defocus curves and to assess the most appropriate criterion for calculating the subjective amplitude of accommodation (AoA) from defocus curves. SETTING: Eye Clinic, School of Life & Health Sciences, Aston University, Birmingham, United Kingdom. METHODS: Defocus curves (from +3.00 diopters [D] to -3.00 D in 0.50 D steps) for 6 possible combinations of randomized or nonrandomized letter sequences and/or lens-presentation order were measured in a random order in 20 presbyopic subjects. Subjective AoA was calculated from the defocus curves by curve fitting using various published criteria, and each was correlated to subjective push-up AoA. Objective AoA was measured for comparison of blur tolerance and pupil size. RESULTS: Randomization of lens-presentation order and/or letter sequences, or lack of, did not affect the measured defocus curves (P>.05, analysis of variance). The range of defocus that maintains highest achievable visual acuity (allowing for variability of repeated measurement) was better correlated to (r = 0.84) and agreed best with ( 0.50 D) subjective push-up AoA than any other relative or absolute acuity criterion used in previous studies. CONCLUSIONS: Nonrandomized letters and lens presentation on their own did not affect subjective AoA measured by defocus curves, although their combination should be avoided. Quantification of subjective AoA from defocus curves should be standardized to the range of defocus that maintains the best achievable visual acuity.

High-power picosecond pulse generation from a gain-switched violet diode laser

Generation of picosecond pulses with a peak power in excess of 7W and a duration of 24ps from a gain-switched InGaN diode laser is demonstrated for the first time.

Multifocal intraocular lens differentiation using defocus curves

Purpose: To determine the most appropriate analysis technique for the differentiation of multifocal intraocular lens (MIOL) designs using defocus curve assessment of visual capability.Methods:Four groups of fifteen subjects were implanted bilaterally with either monofocal intraocular lenses, refractive MIOLs, diffractive MIOLs, or a combination of refractive and diffractive MIOLs. Defocus curves between -5.0D and +1.5D were evaluated using an absolute and relative depth-of-focus method, the direct comparison method and a new 'Area-of-focus' metric. The results were correlated with a subjective perception of near and intermediate vision. Results:Neither depth-of-focus method of analysis were sensitive enough to differentiate between MIOL groups (p>0.05). The direct comparison method indicated that the refractive MIOL group performed better at +1.00, -1.00 and -1.50 D and worse at -3.00, -3.50, -4.00 and -5.00D compared to the diffractive MIOL group (p

Grating and plaid masks indicate linear summation in a contrast gain pool

In human vision, the response to luminance contrast at each small region in the image is controlled by a more global process where suppressive signals are pooled over spatial frequency and orientation bands. But what rules govern summation among stimulus components within the suppressive pool? We addressed this question by extending a pedestal plus pattern mask paradigm to use a stimulus with up to three mask components: a vertical 1 c/deg pedestal, plus pattern masks made from either a grating (orientation = -45°) or a plaid (orientation = ±45°), with component spatial frequency of 3 c/deg. The overall contrast of both types of pattern mask was fixed at 20% (i.e., plaid component contrasts were 10%). We found that both of these masks transformed conventional dipper functions (threshold vs. pedestal contrast with no pattern mask) in exactly the same way: The dipper region was raised and shifted to the right, but the dipper handles superimposed. This equivalence of the two pattern masks indicates that contrast summation between the plaid components was perfectly linear prior to the masking stage. Furthermore, the pattern masks did not drive the detecting mechanism above its detection threshold because they did not abolish facilitation by the pedestal (Foley, 1994). Therefore, the pattern masking could not be attributed to within-channel masking, suggesting that linear summation of contrast signals takes place within a suppressive contrast gain pool. We present a quantitative model of the effects and discuss the implications for neurophysiological models of the process. © 2004 ARVO.