Task-driven saliency using natural statistics (SUN) | Zendy

Matthew H. Tong | Zendy; Christopher Kanan | Zendy; Zhang Li | Zendy; G. Cottrell | Zendy

AI Assistant Blog Pricing

Home ZAIA Blog

Open Access

Task-driven saliency using natural statistics (SUN)

Author(s) -

Matthew H. Tong,

Christopher Kanan,

Zhang Li,

G. Cottrell

Publication year - 2010

Publication title -

journal of vision

Language(s) - English

Resource type - Journals

SCImago Journal Rank - 1.126

H-Index - 113

ISSN - 1534-7362

DOI - 10.1167/9.8.392

Subject(s) - salience (neuroscience) , computer science , probabilistic logic , artificial intelligence , visual search , statistical model , visual attention , eye tracking , eye movement , task (project management) , computational model , computer vision , machine learning , psychology , perception , management , economics , neuroscience

Learning the scales of objects The scale at which objects appear varies significantly so there is no single optimal scale. However the distribution of scales within each object class is distinct. We cluster the percent increase in size of the resized features for each training object using one Gaussian mixture model per class with three clusters each. The three cluster centers found for each class are used to rescale the filters when computing P(C = 1 | F = fz) during testing. For ease of computation, we combine our bottom-up and appearance terms. Recall, this combination is pointwise mutual information.

The content you want is available to Zendy users.

Already have an account? Click here to sign in.

Having issues? You can contact us here

Accelerating Research