Figure 15From: A neural computational model for bottom-up attention with invariant and overcomplete representationInfluences of different parameters in our model with invariant features on a composite stimulus. (a) source image; (b) result of optimized parameters; (c) smaller RF size in layer 1, which fails in detecting the target; (d) bigger RF size in layer 1, which produces a strange result; (e) smaller RF size in layer 2, which does not influence the result significantly; (f) smaller coefficient αwhich produces less strong suppression, but still detects the target as the most salient bars; (g) smaller excitation bandwidth σ ex , which fails in detecting the target.Back to article page