Detection of global salient structure. (a) source images; (b) saliency maps of the attention model with invariant features coding; (c) saliency maps of the attention model without invariant features coding. Results are relatively insensitive to parameters. σex is 2% to 5% of the image size (the larger one between the image width and height). σinh is 4 to 10 times of σex, α∈[3,4]; (d) saliency map of Itti’s attention model; (e) saliency map of a fully connected network; (f) saliency map of a randomly connected network.
Qi et al. BMC Neuroscience 2012 13:145 doi:10.1186/1471-2202-13-145