Showing papers on "Salience (neuroscience) published in 2006"

Selective Attention and Transfer Phenomena in L2 Acquisition: Contingency, Cue Competition, Salience, Interference, Overshadowing, Blocking, and Perceptual Learning

Abstract: If first language is rational in the sense that acquisition produces an end-state model of language that is a proper reflection of input and that optimally prepares speakers for comprehension and production, second language is usually not. This paper considers the apparent irrationalities of L2 acquisition, that is the shortcomings where input fails to become intake. It describes how ‘learned attention’, a key concept in contemporary associative and connectionist theories of animal and human learning, explains these effects. The fragile features of L2 acquisition are those which, however available as a result of frequency, recency, or context, fall short of intake because of one of the factors of contingency, cue competition, salience, interference, overshadowing, blocking, or perceptual learning, which are all shaped by the L1. Each phenomenon is explained within associative learning theory and exemplified in language learning. Paradoxically, the successes of L1 acquisition and the limitations of L2 acquisition both derive from the same basic learning principles.

An Integrated Model of Top-Down and Bottom-Up Attention for Optimizing Detection Speed

Abstract: Integration of goal-driven, top-down attention and image-driven, bottom-up attention is crucial for visual search. Yet, previous research has mostly focused on models that are purely top-down or bottom-up. Here, we propose a new model that combines both. The bottom-up component computes the visual salience of scene locations in different feature maps extracted at multiple spatial scales. The topdown component uses accumulated statistical knowledge of the visual features of the desired search target and background clutter, to optimally tune the bottom-up maps such that target detection speed is maximized. Testing on 750 artificial and natural scenes shows that the model’s predictions are consistent with a large body of available literature on human psychophysics of visual search. These results suggest that our model may provide good approximation of how humans combine bottom-up and top-down cues such as to optimize target detection speed.

It's under control: top-down search strategies can override attentional capture.

Abstract: Bacon and Egeth (1994) proposed that observed instances of attentional capture by feature singletons (e.g., color) were the result of a salience-based strategy adopted by subjects (singleton detection mode) and, thus, were not automatic. They showed that subjects could override capture by adopting strategies based on searching for specific target features (feature search mode). However, Theeuwes (2004) has recently argued that Bacon and Egeth’s results arose from experimental confounds. He elaborated a model in which attentional capture must be expected when salient distractors fall within a spatialwindow of attention. According to Theeuwes’s (2004) model, there exist two essential criteria for examining stimulus-driven capture. First, search latencies cannot increase with display size, since the search must be parallel; second, the salience of the irrelevant distractor must not be compromised by characteristics of the search display. Contrary to the predictions of Theeuwes’s (2004) model, we provide evidence that involuntary capture can be overridden when both of these criteria are met. Our results are consistent with Bacon and Egeth’s proposal.

Multiple Fundamental Frequency Estimation by Summing Harmonic Amplitudes

Abstract: This paper proposes a conceptually simple and computationally efficient fundamental frequency (F0) estimator for polyphonic music signals. The studied class of estimators calculate the salience, or strength, of a F0 candidate as a weighted sum of the amplitudes of its harmonic partials. A mapping from the Fourier spectrum to a “F0 salience spectrum” is found by optimization using generated training material. Based on the resulting function, three different estimators are proposed: a “direct” method, an iterative estimation and cancellation method, and a method that estimates multiple F0s jointly. The latter two performed as well as a considerably more complex reference method. The number of concurrent sounds is estimated along with their F0s.

Top-down modulation of preattentive processing: Testing the recovery account of contingent capture

Abstract: One highly controversial issue with respect to preattentive processing concerns the degree to which the preattentive detection of “singletons” elicits an involuntary shift of spatial attention (i.e., attentional capture) that is immune from top-down modulation. According to the “pure-capture” perspective, preattentive processing drives the allocation of spatial attention in a purely bottom-up manner, in order of relative salience. According to the “contingent-capture” perspective, preattentive processing can produce attentional capture, but such capture is contingent on whether the eliciting stimulus carries a feature property consistent with the current attentional set. Pure-capture proponents have recently argued that the evidence for contingencies in attentional capture actually reflects the rapid disengagement and recovery from capture. Two spatial cueing experiments tested the rapid recovery by measuring (1) compatibility effects associated with irrelevant distractors and (2) inhibition of return to ...

Learning by selection: visual search and object perception in young infants.

Abstract: The authors examined how visual selection mechanisms may relate to developing cognitive functions in infancy. Twenty-two 3-month-old infants were tested in 2 tasks on the same day: perceptual completion and visual search. In the perceptual completion task, infants were habituated to a partly occluded moving rod and subsequently presented with unoccluded broken and complete rod test stimuli. In the visual search task, infants viewed displays in which single targets of varying levels of salience were cast among homogeneous static vertical distractors. Infants whose posthabituation preference indicated unity perception in the completion task provided evidence of a functional visual selective attention mechanism in the search task. The authors discuss the implications of the efficiency of attentional mechanisms for information processing and learning.

Computation of intrinsic perceptual saliency in visual environments, and applications

Abstract: Detection of image salience in a visual display of an image. The image is analyzed at multiple spatial scales and over multiple feature channels to determine the likely salience of different portions of the image. One application for the system is in an advertising context. The detection may be improved by second order statistics, e.g. mean and the standard deviations of different image portions relative to other portions. Different edges may be considered as being extended edges by looking at the edges over multiple spatial scales. One set of feature channels can be optimized for use in moving images, and can detect motion or flicker. The images can be obtained over multiple spectral ranges the user can be instructed about how to maximize the saliency. This can be applied to automatically evaluate and optimize sales or advertisement displays.

Quantitative modelling of perceptual salience at human eye position

Abstract: We investigate the extent to which a simple model of bottom-up attention and salience may be embedded within a broader computational framework, and compared with human eye movement data. We focus on quantifying whether increased simulation realism significantly affects quantitative measures of how well the model may predict where in video clips humans direct their gaze. We hence compare three variants of the model, tested with 15 video clips of natural scenes shown to three observers. We measure model-predicted salience at the locations gazed to by the observers, compared to random locations. The first variant simply processes the raw video clips. The second adds a gaze-contingent foveation filter. The third further attempts to realistically simulate dynamic human vision by embedding the video frames within a larger background, and shifting them to eye position. Our main finding is that increasing simulation realism significantly improves the predictive ability of the model. Better emulating the details of how a visual stimulus is captured by a constantly rotating retina during active vision has a significant positive impact onto quantitative comparisons between model and human behaviour.

Region Enhanced Scale-Invariant Saliency Detection

Abstract: Saliency measures the low-level stimuli to human vision, and serves as an alternative to semantic image understanding. This paper presents a region enhanced scale-invariant saliency detection method. Our method constructs a scale-invariant saliency map from an image, segments the image into regions, and enhances the saliency map with the region information. Compared with previous methods, our method has advantages in providing robust scale-invariant saliency, giving meaningful region information for applications, and eliminating misleading high-contrast edges.

Integration of exogenous input into a dynamic salience map revealed by perturbing attention.

Abstract: Although it is widely accepted that exogenous and voluntary factors jointly determine the locus of attention, the rules governing the integration of these factors are poorly understood. We investigated neural responses in the lateral intraparietal area (LIP) to transient, distracting visual perturbations presented during task performance. Monkeys performed a covert search task in which they discriminated the orientation of a target embedded among distractors, and brief visual perturbations were presented at various moments and locations during task performance. LIP neurons responded to perturbations consisting of the appearance of new objects, as well as to abrupt changes in the color, luminance, or position of existing objects. The LIP response correlated with the bottom-up behavioral effects of different perturbation types. In addition, neurons showed two types of top-down modulations. One modulation was a context-specific multiplicative gain that affected perturbation, target, and distractor activity in a spatially nonspecific manner. Gain was higher in blocks of trials in which perturbations directly marked target location than in blocks in which they invariably appeared opposite the target, thus encoding a behavioral context defined by the statistical contingency between target and perturbation location. A second modulation reflected local competitive interactions with search-related activity, resulting in the converse effect: weaker perturbation-evoked responses if perturbations appeared at the location of the target than if they appeared opposite the target. Thus, LIP encodes an abstract dimension of salience, which is shaped by local and global top-down mechanisms. These interacting mechanisms regulate responsiveness to external input as a function of behavioral context and momentary task demands.

Dopamine, prediction error and associative learning: A model-based account

Abstract: The notion of prediction error has established itself at the heart of formal models of animal learning and current hypotheses of dopamine function. Several interpretations of prediction error have been offered, including the model-free reinforcement learning method known as temporal difference learning (TD), and the important Rescorla–Wagner (RW) learning rule. Here, we present a model-based adaptation of these ideas that provides a good account of empirical data pertaining to dopamine neuron firing patterns and associative learning paradigms such as latent inhibition, Kamin blocking and overshadowing. Our departure from model-free reinforcement learning also offers: 1) a parsimonious distinction between tonic and phasic dopamine functions; 2) a potential generalization of the role of phasic dopamine from valence-dependent “reward” processing to valence-independent “salience” processing; 3) an explanation for the selectivity of certain dopamine manipulations on motivation for distal rewards; and 4) a plau...

Spatial coding in two dimensions.

Abstract: There has been growing interest in exploring human performance for situations in which stimuli and/or responses vary along both horizontal and vertical dimensions. Earlier studies indicated that there is a prevalence of the horizontal dimension over the vertical dimension in the spatial codes that are used for response selection. We review evidence about spatial coding for 2-D stimulus-response sets and accounts that have been proposed for explaining how it takes place. Particular attention is devoted to the relative salience account, which provides the most comprehensive explanation of 2-D spatial coding. We also evaluate the influence of speed of spatial code formation, number of reference frames, and learning on subjects' performance in 2-D tasks.

Effects of presentation speed of a dynamic visualization on the understanding of a mechanical system

Abstract: Effects of Presentation Speed of a Dynamic Visualization on the Understanding of a Mechanical System Sebastian Fischer (s.fischer@iwm-kmrc.de) Knowledge Media Research Center, Konrad-Adenauer-Strasse 40, 72072 T¨ ubingen, Germany Richard K. Lowe (r.k.lowe@curtin.edu.au) Curtin University, GPO Box U1987, Perth 6845, WA, Australia Stephan Schwan (s.schwan@iwm-kmrc.de) Knowledge Media Research Center, Konrad-Adenauer-Strasse 40, 72072 T¨ ubingen, Germany Abstract dynamic visualisation (such as the size, shape, colour, and arrangement of its component entities) and its tem- poral properties (such as playing speed, direction, and continuity) have the potential to affect the relative per- ceptual salience of displayed information. In an experimental study, the role of temporal variation of a realistic animation was examined. The animation of a complex mechanical system, a pendulum clock, was presented in a between subject design at normal or fast speed. Presentation speed was found to affect distribu- tion of attention and understanding of the functional- ity of the clockwork mechanism. Verbal reports in the fast condition contained more statements on the weight, which is a central part of the clocks’ mechanism. When giving a written description of the clock, subjects who saw the fast presentation produced more correct and less false concepts about the key components of the clock- work. For complex subject matter, it seems possible that speed could be used strategically by instructional designers to raise the perceptual salience of thematically relevant aspects of the display, by means of faster or slower presentation speeds. Instructional Uses of Temporal Manipulations Accordingly, research into learning from dynamic vi- sualizations has produced mixed findings. For exam- ple, while Schwan and Riempp (2004), found that user- controllable videos facilitated the learning of knot-tying tasks, no corresponding facilitation effect was found in studies by Lowe (2003) of learning meteorological pre- diction skills from a user-controllable weather map an- imation. A likely reason for the mixed results is that such studies differed in a number of key respects. One important way they differ is in the type of content that is presented, which can range from concrete subject matter such as mechanical devices to highly abstract material such as visualizations of mathematical algorithms. In addition to these content differences, there are also dif- ferences in how literally the specific content is depicted. The issue here is one of the extent to which the visuospa- tial and temporal properties of the depiction resemble those of the real life referent. In terms of visuospatial properties, the appearance of the depicted content can be presented with varying degrees of realism, ranging from extremely realistic portrayals (as with films, videos, and realistically rendered animations) through to highly schematic renditions that embody extensive manipula- tion of the referent (such as animated diagrams). With temporal properties, the behaviour of the depicted con- tent can also be presented with varying degrees of re- alism. At one extreme, the dynamic visualization has a strict moment-to-moment correspondence with the refer- ent situation; at the other, temporal characteristics such as presentation speed and order have been extensively manipulated. This paper focuses upon temporal properties of a dy- namic display and explores the effect of presentation speed on the processing of information in an animated depiction of a complex mechanical device (a pendulum clock). Dynamic visualizations of various types are increas- ingly used in educational resources for presenting sub- ject matter that involves change over time. Intuitively, these representations appear well suited to helping learn- ers understand change phenomena because they provide explicit depiction of the referent situation’s dynamics in- stead of requiring the learner to reconstruct the dynam- ics via mental animation from static graphics. Never- theless, dynamic visualizations have not proven to be a universal panacea for the challenges involved in compre- hending change phenomena. Merely providing change- related information as an analogue dynamic represen- tation does not guarantee that learners will necessarily be able to extract its key aspects (Lowe, 1999, 2003, 2004). Particularly when the subject matter is complex and unfamiliar, learners appear to respond to the conse- quent high level of information processing demands by selectively attending to aspects of the display that have greatest perceptual salience relative to the rest of the display. Unfortunately, such aspects are not necessar- ily those of most relevance to the central theme of the presentation. Both the visuospatial characteristics of a Strict maintenance of the actual temporal properties of a referent system in its depiction can pose process- ing challenges for learners. For example, because ani- mation is a fleeting form of representation (Ainsworth & Van Labeke, 2004), there is potential for learners to miss key information because it is displayed only very briefly or to lose information because it is ‘overwritten’ in working memory by what is presented subsequently (Lowe, 1999). Such failures to extract or retain key in- formation are likely to prejudice the individual’s capacity

Salience and target selection in visual search.

Abstract: Why is it easy to find strawberries and difficult to collect gooseberries or green tomatoes? Why don't we see the tree in the forest but do see the single tree in the garden? Various explanations have been given to account for these phenomena. The present paper is concentrating on salience, a property apparently important for search but not frequently discussed in this context. Salience lets targets stand out and thus controls the selection of items that need to be investigated when a certain target is to be searched for. This proposal was tested in two series of experiments. It was found that salience detection and target identification followed different time courses; even typical “pop-out” targets (bright or dark blobs, gaps) were faster located than identified. The importance of salience in visual search was further investigated by manipulating target salience without affecting stimulus properties that are assumed to be relevant for search. When the salience of individual items was increased, the effe...

Melody Detection in Polyphonic Musical Signals: Exploiting Perceptual Rules, Note Salience, and Melodic Smoothness

Abstract: 80 Computer Music Journal Melody extraction from polyphonic audio is a research area of increasing interest. It has a wide range of applications in various fields, including music information retrieval (MIR, particularly in query-by-humming, where the user hums a tune to search a database of musical audio), automatic melody transcription, performance and expressiveness analysis, extraction of melodic descriptors for music content metadata, and plagiarism detection, to name but a few. This area has become increasingly relevant in recent years, as digital music archives are continuously expanding. The current state of affairs presents new challenges to music librarians and service providers regarding the organization of large-scale music databases and the development of meaningful methods of interaction and retrieval. In this article, we address the problem of melody detection in polyphonic audio following a multistage approach, inspired by principles from perceptual theory and musical practice. Our system comprises three main modules: pitch detection, determination of musical notes (with precise temporal boundaries, pitches, and intensity levels), and identification of melodic notes. The main contribution of this article is in the last module, in which a number of rule-based systems are proposed that attempt to extract the notes that convey the main melodic line among the whole set of detected notes. The system performs satisfactorily in a small database collected by us and in the database created for the ISMIR 2004 melody extraction contest. However, the performance of the algorithm decreased in the MIREX 2005 database. Related Work

System and method for quantifying and mapping visual salience

Abstract: A system for quantifying and mapping visual salience to a visual stimulus, including a processor, software for receiving data indicative of a group of individual's ocular responses to a visual stimulus, software for determining a distribution of visual resources at each of at least two times for each of at least a portion of the individuals. The system further including software for determining and quantifying a group distribution of visual resources at each of the at least two times and software for generating a display of the group's distribution of visual resources to the visual stimulus.

Task impact on the visual attention in subjective image quality assessment

Abstract: Visual attention is a main feature of the Human Visual System (HVS). Knowing and using the mechanisms of the visual attention could help to the improvement of image quality assessment methods. But, which kind of saliency should be taken into account? A free-task visual selective attention or a quality oriented visual selective attention. We recorded and evaluated the discrepancy between these two types of oculomotor behavior. The results show the impact of the viewing task on visual strategy.

Mechanical invariants are implicated in dynamic touch as a function of their salience in the stimulus flow

Abstract: The authors investigated the mechanical basis of length perception through dynamic touch using specially designed rods in which the various moments of mass distribution (mass, static moment, and rotational inertia) were varied independently. In a series of 4 experiments, exploration style and rod orientation were manipulated such that the relative salience of moments of mass distribution varied markedly. Results showed that perceived length was based on the most salient moments. The authors concluded that the notion of salience is crucial for understanding the implication of moments of mass distribution in length perception and that it should play a pivotal role in developing an encompassing theory of dynamic touch.

Target-nontarget similarity modulates stimulus-driven control in visual search

Abstract: The literature contains conflicting results concerning whether an irrelevant featural singleton (an item unique with respect to a feature such as color or brightness) can control attention in a stimulus-driven manner. The present study explores whether target-nontarget similarity influences stimulus-driven shifts of attention to a distractor. An experiment evaluated whether manipulating target-nontarget similarity by varying orientation would modulate distraction by an irrelevant feature (a bright singleton). We found that increasing target-nontarget similarity resulted in a decreased impact of a uniquely bright object on visual search. This method of manipulating the target-nontarget similarity independent of the salience of a distracting feature suggests that the extent to which visual attention is stimulus-driven depends on the target-nontarget similarity.

Attention and salience in associative blocking.

Abstract: The associative learning effect called blocking has previously been found in many cue-competition paradigms where all cues are of equal salience. Previous research by Hall, Mackintosh, Goodall, and dal Martello (1977) found that, in animals, salient cues were less likely to be blocked. Crucially, they also found that when the to-be-blocked cue was highly salient, the blocking cue would lose some control over responding. The present article extends these findings to humans and suggests that shifts in attention can explain the apparent loss of control by the previously learned cue. A connectionist model that implements attentional learning is shown to fit the main trends in the data. Model comparisons suggest that mere forgetting, implemented as weight decay, cannot explain the results.

Interrupting infants’ persisting object representations: an object-based limit?

Abstract: Making sense of the visual world requires keeping track of objects as the same persisting individuals over time and occlusion. Here we implement a new paradigm using 10-month-old infants to explore the processes and representations that support this ability in two ways. First, we demonstrate that persisting object representations can be maintained over brief interruptions from additional independent events – just as a memory of a traffic scene may be maintained through a brief glance in the rearview mirror. Second, we demonstrate that this ability is nevertheless subject to an object-based limit: if an interrupting event involves enough objects (carefully controlling for overall salience), then it will impair the maintenance of other persisting object representations even though it is an independent event. These experiments demonstrate how object representations can be studied via their ‘interruptibility’, and the results are consistent with the idea that infants’ persisting object representations are constructed and maintained by capacity-limited mid-level ‘object-files’.