Limits...
An exponential filter model predicts lightness illusions.

Zeman A, Brooks KR, Ghebreab S - Front Hum Neurosci (2015)

Bottom Line: Conversely, assimilation is where the lightness of the target patch moves toward that of the bounding areas and can be demonstrated in White's effect.Combining two filters together increased the best performance to 23, with asymptotic performance at 24 for an arbitrarily large combination of filter outputs.While normalization improved prediction magnitudes, it only slightly improved overall scores in direction predictions.

View Article: PubMed Central - PubMed

Affiliation: Department of Cognitive Science, ARC Centre of Excellence in Cognition and its Disorders, Macquarie University Sydney, NSW, Australia ; Commonwealth Scientific and Industrial Research Organisation Marsfield, NSW, Australia ; Perception in Action Research Centre, Macquarie University Sydney, NSW, Australia.

ABSTRACT
Lightness, or perceived reflectance of a surface, is influenced by surrounding context. This is demonstrated by the Simultaneous Contrast Illusion (SCI), where a gray patch is perceived lighter against a black background and vice versa. Conversely, assimilation is where the lightness of the target patch moves toward that of the bounding areas and can be demonstrated in White's effect. Blakeslee and McCourt (1999) introduced an oriented difference-of-Gaussian (ODOG) model that is able to account for both contrast and assimilation in a number of lightness illusions and that has been subsequently improved using localized normalization techniques. We introduce a model inspired by image statistics that is based on a family of exponential filters, with kernels spanning across multiple sizes and shapes. We include an optional second stage of normalization based on contrast gain control. Our model was tested on a well-known set of lightness illusions that have previously been used to evaluate ODOG and its variants, and model lightness values were compared with typical human data. We investigate whether predictive success depends on filters of a particular size or shape and whether pooling information across filters can improve performance. The best single filter correctly predicted the direction of lightness effects for 21 out of 27 illusions. Combining two filters together increased the best performance to 23, with asymptotic performance at 24 for an arbitrarily large combination of filter outputs. While normalization improved prediction magnitudes, it only slightly improved overall scores in direction predictions. The prediction performance of 24 out of 27 illusions equals that of the best performing ODOG variant, with greater parsimony. Our model shows that V1-style orientation-selectivity is not necessary to account for lightness illusions and that a low-level model based on image statistics is able to account for a wide range of both contrast and assimilation effects.

No MeSH data available.


Related in: MedlinePlus

Power spectra for images that are unfiltered (left column) and filtered with size = 5 pixels (right column). Top row: 28 natural images. Bottom row: 28 illusory images.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4478851&req=5

Figure 8: Power spectra for images that are unfiltered (left column) and filtered with size = 5 pixels (right column). Top row: 28 natural images. Bottom row: 28 illusory images.

Mentions: The filtering approach we use reshapes contrast distributions toward those that best describe natural images using the exponential filter family. Similarly to Dakin and Bex (2003), we essentially reconstruct an image that represents the most probable naturally occurring source. By redistributing lightness values to more closely reflect the underlying statistical relationships of images within our environment, we can form predictions of perceptual lightness estimates that align with a large array of lightness illusions. Figure 8 illustrates the power spectra for a set of images that are unfiltered (left column) and filtered (right column) using different shape filters that are all of size 5 pixels. The top row illustrates power spectra for 28 natural images. From these graphs we can see that the power spectra for filtered natural images is quite similar to the power spectra for unfiltered natural images. The bottom row shows the power spectra for illusory images. The unfiltered images in the bottom left graph show a flatter power spectrum in the lower SFs than the filtered images in the bottom right graph. By applying these exponential filters, we see that they not only push the power spectra of illusory images toward that of natural images, reflecting the properties of image statistics. Applying these filters also boosts low SF information, hypothesized to be a driving factor in the perception of lightness illusions (Dakin and Bex, 2003). Dakin and Bex (2003) find that low spatial frequencies are primarily responsible for the Craik, Cornsweet, and O'Brien (CCOB) illusion that they study. The LoG filters that they apply boost this information when it is not present. From their results, Dakin and Bex (2003) hypothesize that low SF information may drive many illusions.


An exponential filter model predicts lightness illusions.

Zeman A, Brooks KR, Ghebreab S - Front Hum Neurosci (2015)

Power spectra for images that are unfiltered (left column) and filtered with size = 5 pixels (right column). Top row: 28 natural images. Bottom row: 28 illusory images.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4478851&req=5

Figure 8: Power spectra for images that are unfiltered (left column) and filtered with size = 5 pixels (right column). Top row: 28 natural images. Bottom row: 28 illusory images.
Mentions: The filtering approach we use reshapes contrast distributions toward those that best describe natural images using the exponential filter family. Similarly to Dakin and Bex (2003), we essentially reconstruct an image that represents the most probable naturally occurring source. By redistributing lightness values to more closely reflect the underlying statistical relationships of images within our environment, we can form predictions of perceptual lightness estimates that align with a large array of lightness illusions. Figure 8 illustrates the power spectra for a set of images that are unfiltered (left column) and filtered (right column) using different shape filters that are all of size 5 pixels. The top row illustrates power spectra for 28 natural images. From these graphs we can see that the power spectra for filtered natural images is quite similar to the power spectra for unfiltered natural images. The bottom row shows the power spectra for illusory images. The unfiltered images in the bottom left graph show a flatter power spectrum in the lower SFs than the filtered images in the bottom right graph. By applying these exponential filters, we see that they not only push the power spectra of illusory images toward that of natural images, reflecting the properties of image statistics. Applying these filters also boosts low SF information, hypothesized to be a driving factor in the perception of lightness illusions (Dakin and Bex, 2003). Dakin and Bex (2003) find that low spatial frequencies are primarily responsible for the Craik, Cornsweet, and O'Brien (CCOB) illusion that they study. The LoG filters that they apply boost this information when it is not present. From their results, Dakin and Bex (2003) hypothesize that low SF information may drive many illusions.

Bottom Line: Conversely, assimilation is where the lightness of the target patch moves toward that of the bounding areas and can be demonstrated in White's effect.Combining two filters together increased the best performance to 23, with asymptotic performance at 24 for an arbitrarily large combination of filter outputs.While normalization improved prediction magnitudes, it only slightly improved overall scores in direction predictions.

View Article: PubMed Central - PubMed

Affiliation: Department of Cognitive Science, ARC Centre of Excellence in Cognition and its Disorders, Macquarie University Sydney, NSW, Australia ; Commonwealth Scientific and Industrial Research Organisation Marsfield, NSW, Australia ; Perception in Action Research Centre, Macquarie University Sydney, NSW, Australia.

ABSTRACT
Lightness, or perceived reflectance of a surface, is influenced by surrounding context. This is demonstrated by the Simultaneous Contrast Illusion (SCI), where a gray patch is perceived lighter against a black background and vice versa. Conversely, assimilation is where the lightness of the target patch moves toward that of the bounding areas and can be demonstrated in White's effect. Blakeslee and McCourt (1999) introduced an oriented difference-of-Gaussian (ODOG) model that is able to account for both contrast and assimilation in a number of lightness illusions and that has been subsequently improved using localized normalization techniques. We introduce a model inspired by image statistics that is based on a family of exponential filters, with kernels spanning across multiple sizes and shapes. We include an optional second stage of normalization based on contrast gain control. Our model was tested on a well-known set of lightness illusions that have previously been used to evaluate ODOG and its variants, and model lightness values were compared with typical human data. We investigate whether predictive success depends on filters of a particular size or shape and whether pooling information across filters can improve performance. The best single filter correctly predicted the direction of lightness effects for 21 out of 27 illusions. Combining two filters together increased the best performance to 23, with asymptotic performance at 24 for an arbitrarily large combination of filter outputs. While normalization improved prediction magnitudes, it only slightly improved overall scores in direction predictions. The prediction performance of 24 out of 27 illusions equals that of the best performing ODOG variant, with greater parsimony. Our model shows that V1-style orientation-selectivity is not necessary to account for lightness illusions and that a low-level model based on image statistics is able to account for a wide range of both contrast and assimilation effects.

No MeSH data available.


Related in: MedlinePlus