|
This page presents results of an experiment where the proposed
bottom-up
discriminant saliency detector is used to predict human eye fixations,
for natural images and free viewing conditions. Quantitative comparisons with
other methods are presented in the
related publications
(see, e.g., Gao, Mahadevan, and Vasconcelos, Journal of Vision, 2008).
Discriminant saliency maps and human eye fixation maps
These are some example images (top), their discriminant saliency maps (middle)
and "pseudo human saliency maps" generated from the corresponding human eye
fixations (bottom).
|
