nips nips2006 nips2006-86 nips2006-86-reference knowledge-graph by maker-knowledge-mining

86 nips-2006-Graph-Based Visual Saliency

Source: pdf

Author: Jonathan Harel, Christof Koch, Pietro Perona

Abstract: A new bottom-up visual saliency model, Graph-Based Visual Saliency (GBVS), is proposed. It consists of two steps: rst forming activation maps on certain feature channels, and then normalizing them in a way which highlights conspicuity and admits combination with other maps. The model is simple, and biologically plausible insofar as it is naturally parallelized. This model powerfully predicts human xations on 749 variations of 108 natural images, achieving 98% of the ROC area of a human-based control, whereas the classical algorithms of Itti & Koch ([2], [3], [4]) achieve only 84%. 1

reference text

[1] W. Einhäuser, W. Kruse, K.P. Hoffmann, & P. König

[2] L. Itti, C. Koch, & E. Niebur

[3] L. Itti & C. Koch

[4] L. Itti, & P. Baldi

[5] N. Bruce & J. Tsotsos

[6] L.F. Costa

[7] G. Boccignone, & M. Ferraro

[8] D. Brockmann, T. Geisel

[9] D. Parkhurst, K. Law, & E. Niebur

[10] D.K. Lee, L. Itti, C. Koch, & J. Braun

[11] L. Itti, J. Braun, D.K. Lee, & C. Koch

[12] W. Einhäuser & P. König,

[13] U. Rutishauser, D. Walther, C. Koch, & P. Perona

[14] B.W. Tatler, R.J. Baddeley, & I.D. Gilchrist

[15] J. Malik & P. Perona