Effectiveness of a Kalman filter model constrained by efficient coding in explaining the coexistence of repulsive and attractive perceptual biases.

Human perceptual decisions can be repelled away from (repulsive adaptation) or attracted towards recent visual experience (attractive serial dependence). It is currently unclear whether and how these repulsive and attractive biases interact during visual processing and what computational principles underlie these history dependencies. Here we disentangle repulsive and attractive biases by exploring their respective timescales. We find that perceptual decisions are concurrently attracted towards the short-term perceptual history and repelled from stimuli experienced up to minutes into the past. The temporal pattern of short-term attraction and long-term repulsion cannot be captured by an ideal Bayesian observer model alone. Instead, it is well captured by an ideal observer model with efficient encoding and Bayesian decoding of visual information in a slowly changing environment. Concurrent attractive and repulsive history biases in perceptual decisions may thus be the consequence of the need for visual processing to simultaneously satisfy constraints of efficiency and stability.

Human perceptual decisions can be repelled away from (repulsive adaptation) or attracted towards recent visual experience (attractive serial dependence). It is currently unclear whether and how these repulsive and attractive biases interact during visual processing and what computational principles underlie these history dependencies. Here we disentangle repulsive and attractive biases by exploring their respective timescales. We find that perceptual decisions are concurrently attracted towards the short-term perceptual history and repelled from stimuli experienced up to minutes into the past. The temporal pattern of short-term attraction and long-term repulsion cannot be captured by an ideal Bayesian observer model alone. Instead, it is well captured by an ideal observer model with efficient encoding and Bayesian decoding of visual information in a slowly changing environment. Concurrent attractive and repulsive history biases in perceptual decisions may thus be the consequence of the need for visual processing to simultaneously satisfy constraints of efficiency and stability.

Perception of gravity can be assessed by measuring the subjective visual vertical (SVV), the visually indicated spatial direction that appears earth-vertical to an observer. When the SVV is assessed in darkness while the observer is roll-tilted, it shows substantial biases. At tilts larger than 45°, the bias is attractive, that is, the visual indicator appears vertical when rotated toward the observer. At smaller tilts, however, a repulsive bias is observed. The attractive bias has been explained within the Bayesian framework as the effect of a prior for upright posture. The repulsive bias has so far been considered anti-Bayesian, suboptimal, or as the result of uncompensated ocular counterroll. Here we show that both biases can be explained within a purely Bayesian model. More specifically, the repulsive bias at small roll-tilts is a consequence of the known tilt-dependent variability of the SVV, which is hypothesized to reflect different levels of sensory noise of the otolith organs. We thus provide a solution to a century-old question of why there is a repulsive bias in vertical perception.

BackgroundSequential effects of environmental stimuli are ubiquitous in most behavioral tasks involving magnitude estimation, memory, decision making, and emotion. The human visual system exploits continuity in the visual environment, which induces two contrasting perceptual phenomena shaping visual perception. Previous work reported that perceptual estimation of a stimulus may be influenced either by attractive serial dependencies or repulsive aftereffects, with a number of experimental variables suggested as factors determining the direction and magnitude of sequential effects. Recent studies have theorized that these two effects concurrently arise in perceptual processing, but empirical evidence that directly supports this hypothesis is lacking, and it remains unclear whether and how attractive and repulsive sequential effects interact in a trial. Here we show that the two effects concurrently modulate estimation behavior in a typical sequence of perceptual tasks.ResultsWe first demonstrate that observers’ estimation error as a function of both the previous stimulus and response cannot be fully described by either attractive or repulsive bias but is instead well captured by a summation of repulsion from the previous stimulus and attraction toward the previous response. We then reveal that the repulsive bias is centered on the observer’s sensory encoding of the previous stimulus, which is again repelled away from its own preceding trial, whereas the attractive bias is centered precisely on the previous response, which is the observer’s best prediction about the incoming stimuli.ConclusionsOur findings provide strong evidence that sensory encoding is shaped by dynamic tuning of the system to the past stimuli, inducing repulsive aftereffects, and followed by inference incorporating the prediction from the past estimation, leading to attractive serial dependence.

Bayesian observer models have been quite successful in accounting for perceptual behavior. However, it is a common challenge to specify the two fundamental components of a Bayesian model, the prior distribution and the likelihood function, because they are formally unconstrained. We argue that a perceptual system that is adapted to the statistical structure of its environment naturally imposes constraints on its corresponding Bayesian model description. In particular, we assume the prior to reflect the stimulus distribution and the likelihood to be constrained by a sensory representation that is efficient. We show that these assumptions lead to an observer model that makes two counter-intuitive predictions: First, perceptual biases can be repulsive (i.e. biased away from the peak of the prior), which is in stark contrast to the traditional Bayesian view. Second, the model predicts that neural and stimulus noise are differentially affecting perceptual bias, with larger neural noise leading to an increase in repulsive bias while larger stimulus noise leading to a decrease. We tested our model against reported experimental data regarding two perceptual variables for which the natural stimulus statistics are known (orientation and spatial frequency of visual stimuli). We found that the model predicts the reported repulsive biases from the cardinal orientations and low spatial frequencies, respectively. Furthermore, it also accounts for the observed increase in bias with increasing levels of neural noise, as well as the relative attractive bias when comparing stimuli with high versus low stimulus noise. The model is capable of making quantitative predictions up to a scaling factor for any perceptual variable for which the stimulus statistics are known. Our results suggest that efficient coding provides a powerful constraint in specifying Bayesian observer models, and leads to successful predictions of perceptual effects that have been considered incompatible with the Bayesian framework. Meeting abstract presented at VSS 2014

Attractiveness bias in observational data

Despite the old adage not to ‘judge a book by its cover’, facial cues often guide first impressions and these first impressions guide our decisions. Literature suggests there are valid facial cues that assist us in assessing someone’s health or intelligence, but such cues are overshadowed by an ‘attractiveness halo’ whereby desirable attributions are preferentially ascribed to attractive people. The impact of the attractiveness halo effect on perceptions of academic performance in the classroom is concerning as this has shown to influence students’ future performance. We investigated the limiting effects of the attractiveness halo on perceptions of actual academic performance in faces of 100 university students. Given the ambiguity and various perspectives on the definition of intelligence and the growing consensus on the importance of conscientiousness over intelligence in predicting actual academic performance, we also investigated whether perceived conscientiousness was a more accurate predictor of academic performance than perceived intelligence. Perceived conscientiousness was found to be a better predictor of actual academic performance when compared to perceived intelligence and perceived academic performance, and accuracy was improved when controlling for the influence of attractiveness on judgments. These findings emphasize the misleading effect of attractiveness on the accuracy of first impressions of competence, which can have serious consequences in areas such as education and hiring. The findings also have implications for future research investigating impression accuracy based on facial stimuli.

Vision can be considered as a process of probabilistic inference. In a Bayesian framework, perceptual estimates from sensory information are combined with prior knowledge, with a stronger influence of the prior when the sensory evidence is less certain. Here, we explored the behavioral and neural consequences of manipulating stimulus certainty in the context of orientation processing. First, we asked participants to judge whether a stimulus was oriented closer to vertical or the clockwise primary oblique (45°) for two stimulus types (spatially filtered noise textures and sinusoidal gratings) and three manipulations of certainty (orientation bandwidth, contrast, and duration). We found that participants consistently had a bias toward reporting orientation as closer to 45° during conditions of high certainty and that this bias was reduced when sensory evidence was less certain. Second, we measured event-related fMRI BOLD responses in human primary visual cortex (V1) and manipulated certainty via stimulus contrast (100% vs 3%). We then trained a multivariate classifier on the pattern of responses in V1 to cardinal and primary oblique orientations. We found that the classifier showed a bias toward classifying orientation as oblique at high contrast but categorized a wider range of orientations as cardinal for low-contrast stimuli. Orientation classification based on data from V1 thus paralleled the perceptual biases revealed through the behavioral experiments. This pattern of bias cannot be explained simply by a prior for cardinal orientations.SIGNIFICANCE STATEMENT Our perception of the world around us is biased through prior expectations rather than necessarily reflecting the true state of our environment. Here, we investigate biases in the visual processing of spatial orientation to understand how prior expectations and current sensory information interact to generate a percept. By degrading visual input in various ways, we are able to quantify the extent to which prior experience affects both perceptual judgments and neural responses in the human visual system. We observe systematic biases in the perception of orientation that correlate with the pattern of activity in the primary visual cortex of the human brain. These results indicate that prior expectations influence neural processing right from the earliest stage of the cortical hierarchy.

Perceptions of source credibility may play a role in major societal challenges like political polarization and the spread of misinformation as citizens disagree over which sources of political information are credible and sometimes trust untrustworthy sources. Cognitive scientists have developed Bayesian Network models of how people integrate perceptions of source credibility when learning from information provided by sources, but these models do not involve the crucial source characteristic in politics: bias. Biased sources make claims that align with a particular political agenda, whether or not they are true. We present a novel Bayesian Network model which integrates perceptions of a source's bias as well as their expertise. We demonstrate the model's validity for predicting how people will update beliefs and perceptions of bias and expertise in response to testimony across two studies, the second being a preregistered conceptual replication and extension of the first.

Human smooth pursuit eye movements and motion perception behave similarly when observers track and judge the motion of simple objects, such as dots. But moving objects in our natural environment are complex and contain internal motion. We ask how pursuit and perception integrate the motion of objects with motion that is internal to the object. Observers (n = 20) tracked a moving random-dot kinematogram with their eyes and reported the object's perceived direction. Objects moved horizontally with vertical shifts of 0, ±3, ±6, or ±9° and contained internal dots that were static or moved ±90° up/down. Results show that whereas pursuit direction was consistently biased in the direction of the internal dot motion, perceptual biases differed between observers. Interestingly, the perceptual bias was related to the magnitude of the pursuit bias (r = 0.75): perceptual and pursuit biases were directionally aligned in observers that showed a large pursuit bias, but went in opposite directions in observers with a smaller pursuit bias. Dissociations between perception and pursuit might reflect different functional demands of the two systems. Pursuit integrates all available motion signals in order to maximize the ability to monitor and collect information from the whole scene. Perception needs to recognize and classify visual information, thus segregating the target from its context. Ambiguity in whether internal motion is part of the scene or contributes to object motion might have resulted in individual differences in perception. The perception-pursuit correlation suggests shared early-stage motion processing or perception-pursuit interactions.

Human smooth pursuit eye movements and motion perception behave similarly when observers track and judge the motion of simple objects, such as dots. But moving objects in our natural environment are complex and contain internal motion. We ask how pursuit and perception integrate the motion of objects with motion that is internal to the object. Observers (n = 20) tracked a moving random-dot kinematogram with their eyes and reported the object’s perceived direction. Objects moved horizontally with vertical shifts of 0, ±3, ±6, or ±9° and contained internal dots that were static or moved ±90° up/down. Results show that whereas pursuit direction was consistently biased in the direction of the internal dot motion, perceptual biases differed between observers. Interestingly, the perceptual bias was related to the magnitude of the pursuit bias (r = 0.75): perceptual and pursuit biases were directionally aligned in observers that showed a large pursuit bias, but went in opposite directions in observers with a smaller pursuit bias. Dissociations between perception and pursuit might reflect different functional demands of the two systems. Pursuit integrates all available motion signals in order to maximize the ability to monitor and collect information from the whole scene. Perception needs to recognize and classify visual information, thus segregating the target from its context. Ambiguity in whether internal motion is part of the scene or contributes to object motion might have resulted in individual differences in perception. The perception-pursuit correlation suggests shared early-stage motion processing or perception-pursuit interactions.

In general, viewers are more attracted to local features in images at a shorter viewing distance and to global features in images at a longer viewing distance. However, numerical analysis of the effect of viewing distance on human texture perception and how the perception of global and local changes under certain conditions are still undetermined. In this paper, we present statistical prediction of the relationship between the domination ratio of global and local features and the viewing distances under the control of several factors, using the logistic regression model. We synthesized textures by separately controlling global and local textural features using a texture model based on mathematical morphology, namely the primitive, grain, and point configuration texture model. Visual sensory tests were carried out on 80 subjects during two sets of experiments. The collected data were statistically analyzed using logistic regression and Akaike information criteria. Besides the main factor of viewing distance, the factors including gender, changing the order of viewing positions, and prior knowledge were also shown quantitatively to have significant influence on human texture perception. Our results showed that (1) local features of a texture were more attractive to females than males, (2) the first impression might have affected subsequent decisions in texture perception, and (3) subjects who had prior knowledge (supervised) were more sensitive to the changes in global and local dominance. (4) Regarding the interactions of the factors, prior knowledge reduced the effects of individual differences and perception condition differences on human texture perception. This study is dedicated to the construction of numerical relationships between viewing distance and human texture perception as well as to cognitive investigation of biases in global and local perceptions.

Background clutter characterization in infrared imagery has become an actively researched field, and several clutter models have been reported. These models attempt to evaluate the target detection and recognition probabilities that are characteristic of a certain scene when specific target and human visual perception features are known. The prior knowledge assumed and required by these models is a severe limitation. Furthermore, the attempt to model subjective and intricate mechanisms such as human perception with general mathematical formulas is controversial. In this paper, we introduce the idea of adaptive models that are dynamically derived from a set of examples by a supervised learning mechanism based on genetic programming foundations. A set of characteristic scene and target features with a demonstrated influence on the human visual perception mechanism is first extracted from the original images. Then, the correlations between these features and detection performance results obtained by visual observer tests on the same set of images are captured into models by a learning algorithm. The effectiveness of the adaptive modeling principle is discussed in the final part of the paper.

Background clutter characterization in IR imagery has become an actively researched field and several clutter models have been reported. These models attempt to evaluate the target detection/recognition probabilities that are characteristic to a certain scene when specific target and human visual perception features are known. The prior knowledge assumed and required by these models is a severe limitation. Furthermore, the attempt to model subjective and intricate mechanisms such as human perception with simple mathematical formulae is controversial. In this paper, we introduce the idea of adaptive models that are dynamically derived from a set of examples by a supervised evolutionary learning scheme. A set of characteristic scene and target features with a demonstrated influence on the human visual perception mechanism is first extracted from the original images. Then, the correlation between these features and the results obtained by visual observer tests on the same set of images are captured into a model by the learning scheme. The effectiveness of the adaptive modeling principle is discussed in the final part of the paper.

Decision letter: A Bayesian and efficient observer model explains concurrent attractive and repulsive history biases in visual perception