Limits...
The mechanisms of feature inheritance as predicted by a systems-level model of visual attention and decision making.

Hamker FH - Adv Cogn Psychol (2008)

Bottom Line: We find that the presence of feedback loops alone is sufficient to account for feature inheritance.Although our simulations do not cover all experimental variations and focus only on the general principle, our result appears of specific interest since the model was designed for a completely different purpose than to explain feature inheritance.We suggest that feedback is an important property in visual perception and provide a description of its mechanism and its role in perception.

View Article: PubMed Central - PubMed

Affiliation: Department of Psychology, Westf.-Wilhelms-Universität Münster, Germany.

ABSTRACT
Feature inheritance provides evidence that properties of an invisible target stimulus can be attached to a following mask. We apply a systemslevel model of attention and decision making to explore the influence of memory and feedback connections in feature inheritance. We find that the presence of feedback loops alone is sufficient to account for feature inheritance. Although our simulations do not cover all experimental variations and focus only on the general principle, our result appears of specific interest since the model was designed for a completely different purpose than to explain feature inheritance. We suggest that feedback is an important property in visual perception and provide a description of its mechanism and its role in perception.

No MeSH data available.


Population activity in IT from target onset to mask offset in three different						model conditions, sensory-driven perception without feedback, sensory driven						perception with feedback and hypothesis-driven feedback with the memorizing						a target template at 180-200 ms after target onset. The numbers on the left						indicate the orientation offset of the target stimulus with respect to the						mask.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2864986&req=5

Figure 7: Population activity in IT from target onset to mask offset in three different model conditions, sensory-driven perception without feedback, sensory driven perception with feedback and hypothesis-driven feedback with the memorizing a target template at 180-200 ms after target onset. The numbers on the left indicate the orientation offset of the target stimulus with respect to the mask.

Mentions: The present results suggest that feature inheritance requires hypothesis-driven perception (active hypothesis testing) where memory permanently distorts the response in IT. The effect also occurs on the level of V4 but to a lesser degree. However, we did not look at the properties of the second approach to the perception of masked visual stimuli, in which the decaying iconic trace feeds the perceptual decision. A sustained distortion of the population response might not be necessary, if we consider that a perceptual choice is made by the accumulation of evidence. Thus, we fed the evidence for a tilted and non-tilted neural response into a model of decision making and determined the response and time of decision (Fig. 7). The perception of a tilt is an indicator for feature inheritance. No tilt either reflects complete masking or shine through. In the sensory-driven perception without feedback no tilt of the mask has been detected. In the sensory-driven perception with feedback, however, the model responds the perception of a tilt for an orientation difference of 15°-30°. The model of hypothesis-driven perception with memory encoding between 180-200 ms and the one with memory encoding between 140-160 ms (not shown) respond almost equal in decision and response time than the model of sensory-driven perception with feedback (passive hypothesis testing). If the memory encoding occurs earlier in time (100-120 ms), the model predicts the perception of a tilt from an orientation offset of 10°-45°. The difference between the two models of sensory-driven perception has not been obtained by a clever adjustment of the decision threshold. For all orientations, in the model without feedback the accumulated evidence for a tilted grating was never close to the threshold. Thus, feedback appears necessary and sufficient for feature inheritance to occur, of course, depending on the timing and similarity of target and mask.


The mechanisms of feature inheritance as predicted by a systems-level model of visual attention and decision making.

Hamker FH - Adv Cogn Psychol (2008)

Population activity in IT from target onset to mask offset in three different						model conditions, sensory-driven perception without feedback, sensory driven						perception with feedback and hypothesis-driven feedback with the memorizing						a target template at 180-200 ms after target onset. The numbers on the left						indicate the orientation offset of the target stimulus with respect to the						mask.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2864986&req=5

Figure 7: Population activity in IT from target onset to mask offset in three different model conditions, sensory-driven perception without feedback, sensory driven perception with feedback and hypothesis-driven feedback with the memorizing a target template at 180-200 ms after target onset. The numbers on the left indicate the orientation offset of the target stimulus with respect to the mask.
Mentions: The present results suggest that feature inheritance requires hypothesis-driven perception (active hypothesis testing) where memory permanently distorts the response in IT. The effect also occurs on the level of V4 but to a lesser degree. However, we did not look at the properties of the second approach to the perception of masked visual stimuli, in which the decaying iconic trace feeds the perceptual decision. A sustained distortion of the population response might not be necessary, if we consider that a perceptual choice is made by the accumulation of evidence. Thus, we fed the evidence for a tilted and non-tilted neural response into a model of decision making and determined the response and time of decision (Fig. 7). The perception of a tilt is an indicator for feature inheritance. No tilt either reflects complete masking or shine through. In the sensory-driven perception without feedback no tilt of the mask has been detected. In the sensory-driven perception with feedback, however, the model responds the perception of a tilt for an orientation difference of 15°-30°. The model of hypothesis-driven perception with memory encoding between 180-200 ms and the one with memory encoding between 140-160 ms (not shown) respond almost equal in decision and response time than the model of sensory-driven perception with feedback (passive hypothesis testing). If the memory encoding occurs earlier in time (100-120 ms), the model predicts the perception of a tilt from an orientation offset of 10°-45°. The difference between the two models of sensory-driven perception has not been obtained by a clever adjustment of the decision threshold. For all orientations, in the model without feedback the accumulated evidence for a tilted grating was never close to the threshold. Thus, feedback appears necessary and sufficient for feature inheritance to occur, of course, depending on the timing and similarity of target and mask.

Bottom Line: We find that the presence of feedback loops alone is sufficient to account for feature inheritance.Although our simulations do not cover all experimental variations and focus only on the general principle, our result appears of specific interest since the model was designed for a completely different purpose than to explain feature inheritance.We suggest that feedback is an important property in visual perception and provide a description of its mechanism and its role in perception.

View Article: PubMed Central - PubMed

Affiliation: Department of Psychology, Westf.-Wilhelms-Universität Münster, Germany.

ABSTRACT
Feature inheritance provides evidence that properties of an invisible target stimulus can be attached to a following mask. We apply a systemslevel model of attention and decision making to explore the influence of memory and feedback connections in feature inheritance. We find that the presence of feedback loops alone is sufficient to account for feature inheritance. Although our simulations do not cover all experimental variations and focus only on the general principle, our result appears of specific interest since the model was designed for a completely different purpose than to explain feature inheritance. We suggest that feedback is an important property in visual perception and provide a description of its mechanism and its role in perception.

No MeSH data available.