Computational modelling of visual attention

@article{Itti2001ComputationalMO,
  title={Computational modelling of visual attention},
  author={Laurent Itti and Christof Koch},
  journal={Nature Reviews Neuroscience},
  year={2001},
  volume={2},
  pages={194-203}
}
  • L. Itti, C. Koch
  • Published 2001
  • Psychology, Medicine
  • Nature Reviews Neuroscience
Five important trends have emerged from recent work on computational models of focal visual attention that emphasize the bottom-up, image-based control of attentional deployment. First, the perceptual saliency of stimuli critically depends on the surrounding context. Second, a unique 'saliency map' that topographically encodes for stimulus conspicuity over the visual scene has proved to be an efficient and plausible bottom-up control strategy. Third, inhibition of return, the process by which… Expand

Paper Mentions

Interventional Clinical Trial
Cognitive impairment in Parkinson's disease (PD) has far-reaching effects on both motor outcomes and quality of life in PD patients. Furthermore, deep brain stimulation (DBS), now… Expand
ConditionsParkinson Disease
InterventionBehavioral, Procedure
CHAPTER 65 – Specifying the Components of Attention in a Visual Search Task
Although commonly treated as a unitary process, attention is more likely a collection of task-related but separable operations. Three components of attention (set, selection, and movement) areExpand
A Focus on Selection for Fixation
A computational explanation of how visual attention, interpretation of visual stimuli, and eye movements combine to produce visual behavior, seems elusive. Here, we focus on one component: howExpand
Interactions of visual attention and object recognition : computational modeling, algorithms, and psychophysics.
TLDR
A new model of bottom-up salient region selection is developed, which estimates the approximate extent of attended proto-objects in a biologically realistic manner and shows that attentional grouping based on bottom- up processes enables successive learning and recognition of multiple objects in cluttered natural scenes. Expand
Neural Mechanisms of Saliency, Attention, and Orienting
TLDR
Evidence for mechanisms in the primate brain that integrate bottom-up saliency with internal goals for flexible orienting towards behaviorally relevant stimuli is reviewed and novel issues and insights are raised that challenge current views about the neural basis of saliency, attention, and orienting. Expand
A Computational Model of Saliency Map Read-Out during Visual Search
TLDR
A new computational model for the inhibition of return is proposed, which is able to examine priority or saliency map in a manner consistent with psychophysical findings and can be considered as a neural implementation of the episodic theory of attention. Expand
Neural mechanisms of bottom-up selection during visual search
  • K. Thompson
  • Psychology
  • 2001 Conference Proceedings of the 23rd Annual International Conference of the IEEE Engineering in Medicine and Biology Society
  • 2001
Models of attention and saccade target selection propose that within the brain there is a topographic map of visual salience that selects, through a winner-take-all mechanism, locations for furtherExpand
Visual Saliency Computations: Mechanisms, Constraints, and the Effect of Feedback
TLDR
It is found that saliency signals are rapidly generated through lateral excitation and inhibition in successive layers of neural populations selective to a single feature, and these signals can be improved by feedback from a higher cortical area that represents a saliency map. Expand
Saliency Processing in the Human Brain
The human visual system is exposed to a vast amount of information every second; a filter is needed to highlight relevant elements. Attention to visual stimuli can be guided by top-down searchExpand
A Dynamic Neural Field Approach to the Covert and Overt Deployment of Spatial Attention
TLDR
This paper identifies and proposes a possible interaction of four elementary mechanisms for selecting the next locus of attention, memorizing the previously attended locations, anticipating the consequences of eye movements and integrating top-down information in order to perform a visual search task with saccadic eye movements. Expand
Bayesian Modeling of Visual Attention
TLDR
This paper proposes a Bayesian approach that explains the optimal integration of top-down cues and bottom-up cues and demonstrates that the proposed visual saliency effectively predicts human gaze in free-viewing of natural scenes. Expand
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 166 REFERENCES
A saliency-based search mechanism for overt and covert shifts of visual attention
TLDR
A detailed computer implementation of a saliency map scheme is described, focusing on the problem of combining information across modalities, here orientation, intensity and color information, in a purely stimulus-driven manner, which is applied to common psychophysical stimuli as well as to a very demanding visual search task. Expand
Modeling Visual Attention via Selective Tuning
TLDR
This model is a hypothesis for primate visual attention, but it also outperforms existing computational solutions for attention in machine vision and is highly appropriate to solving the problem in a robot vision system. Expand
Shifts in selective visual attention: towards the underlying neural circuitry.
TLDR
This study addresses the question of how simple networks of neuron-like elements can account for a variety of phenomena associated with this shift of selective visual attention and suggests a possible role for the extensive back-projection from the visual cortex to the LGN. Expand
A physiological correlate of the 'spotlight' of visual attention
TLDR
A physiological basis for the effects of spatially directed visual attention is identified and retinotopic mapping of attention-related activation was found in primary visual cortex, as well as in dorsomedial and ventral occipital visual areas previously implicated in processing the attended target features. Expand
Feature-based attention influences motion processing gain in macaque visual cortex
TLDR
Non-spatial, feature-based attentional modulation of visual motion processing is demonstrated, and it is shown that attention increases the gain of direction-selective neurons in visual cortical area MT without narrowing the direction-tuning curves. Expand
fMRI evidence for objects as the units of attentional selection
TLDR
Functional magnetic resonance imaging (fMRI) is used to test key predictions of the object-based theory, which proposes that pre-attentive mechanisms segment the visual array into discrete objects, groups, or surfaces, which serve as targets for visual attention. Expand
Feature combination strategies for saliency-based visual attention systems
  • L. Itti, C. Koch
  • Mathematics, Computer Science
  • J. Electronic Imaging
  • 2001
TLDR
Four combination strategies are compared using three databases of natural color images and it is found that strategy (4) and its simplified, computationally efficient approximation yielded significantly better performance than (1), with up to fourfold improvement, while preserving generality. Expand
Attentional modulation of visual motion processing in cortical areas MT and MST
TLDR
It is reported that the responses of direction-selective neurons in monkey visual cortex are greatly influenced by attention, and that this modulation occurs as early in the cortical hierarchy as the level of the middle temporal visual area (MT). Expand
Vision outside the focus of attention
TLDR
It is found that a feature gradient can be detected without measurable impairment of performance even while a concurrent form-recognition task is carried out, in spite of the fact that the form- Recognition task engages focal attention and thus removes attentive resources from the vicinity of the feature gradient. Expand
The representation of visual salience in monkey parietal cortex
TLDR
The results show that under ordinary circumstances the entire visual world is only weakly represented in LIP, with only the most salient or behaviourally relevant objects being strongly represented. Expand
...
1
2
3
4
5
...