Visual saliency: a biologically plausible contourlet-like frequency domain approach

详细信息    查看全文

• 作者：Peng Bian (1)
  Liming Zhang (1)
  
• 关键词：Visual saliency ; Attention selection ; Saliency map ; Divisive normalization
• 刊名：Cognitive Neurodynamics
• 出版年：2010
• 出版时间：September 2010
• 年：2010
• 卷：4
• 期：3
• 页码：189-198
• 全文大小：2107KB
• 参考文献：1. Atick JJ, Redlich AN (1992) What does the retina know about natural scenes? Neural Comput 4:196鈥?10 CrossRef
  2. Barlow HB (1961) The coding of sensory messages. In: Thorpe WH, Zangwill OL (eds) Current problems in animal behavior. Cambridge University Press, Cambridge, pp 331鈥?60
  3. Bruce ND, Tsotsos JK (2006) Saliency based on information maximization. Adv NIPS 18:155鈥?62
  4. Cavanaugh JR, Bair W, Movshon JA (1997) Orientation-selective setting of contrast gain by the surrounds of macaque striate cortex neurons. Neurosci Abstr 23:227.2
  5. Cavanaugh JR, Bair W, Movshon JA (2002) Selectivity and spatial distribution of signals from the receptive field surround in macaque V1 neurons. J Neurosci 88:2547鈥?556
  6. Do MN, Vetterli M (2005) The contourlet transform: an efficient directional multiresolution image representation. Off J Eur Union Inf Not 49:2091鈥?106
  7. Field DJ (1987) Relations between the statistics of natural images and the response profiles of cortical cells. J Opt Soc Am 4:2379鈥?394 CrossRef
  8. Gao D, Mahadevan V, Vasconcelos N (2008) The discriminant center-surround hypothesis for bottom-up saliency. Adv NIPS 20:497鈥?04
  9. Guo CL, Zhang LM (2010) A novel multiresolution spatio temporal saliency detection model and its applications in image and video compression. IEEE Trans IP 19(1):185鈥?98
  10. Guo CL, Ma Q, Zhang LM. (2008). Spatio-temporal saliency detection using phase spectrum of quaternion Fourier transform. In: Proceedings of CVPR 2008
  11. Harel J, Koch C, Perona P (2007) Graph-based visual saliency. Adv NIPS 19:545鈥?52
  12. Hou X, Zhang L (2007) Saliency detection: a spectral residual approach. In: Proceedings of CVPR 2007
  13. Itti L, Koch C, Niebur E (1998) A model of saliency-based visual attention for rapid scene analysis. IEEE Trans PAMI 20(11):1254鈥?259
  14. Itti L, Braun J, Lee DK, Koch C (1999) Attentional modulation of human pattern discrimination psychophysics reproduced by a quantitative model. Neural Netw 19:143鈥?439
  15. Li Z, Dayan P (2006) Pre-attentive visual selection. Neural Netw 19:143鈥?439
  16. Lu ZK, Lin WS, Yang XK, Ong E, Yao S (2005) Modeling visual attention鈥檚 modulatory after effects on visual sensitivity and quality evaluation. IEEE Trans IP 14(11):1928鈥?942
  17. Navon D (1997) Forest before trees鈥攑recedence of global features in visual-perception. Cogn Psychol 9:353鈥?83 CrossRef
  18. Nothdurft HC (2000) Salience from feature contrast: variations with texture density. Vision Res 40(23):3181鈥?200 CrossRef
  19. Ruderman D (1994) The statistics of natural images. Netw Comput Neural Syst 5(4):517鈥?48 CrossRef
  20. Rutishauser U, Walther D, Koch C, Perona P (2004) Is bottom-up attention useful for object recognition? In: Proceedings of CVPR 2004
  21. Simoncelli EP, Schwartz O (1999) Modeling surround suppression in V1 neurons with a statistically-derived normalization model. Adv NIPS 11:153鈥?59
  22. Tatler BW, Baddeley RJ, Gilchrist ID (2005) Visual correlates of fixation selection: effects of scale and time. Vision Res 45:643鈥?59 CrossRef
  23. Walther D, Koch C (2006) Modeling attention to salient proto-objects. Neural Netw 19:1395鈥?407 CrossRef
  
• 作者单位：Peng Bian (1)
  Liming Zhang (1)
  
  1. Department of Electronic Engineering, Fudan University, Shanghai, 200433, China
  

文摘

In this paper we propose a fast frequency domain saliency detection method that is also biologically plausible, referred to as frequency domain divisive normalization (FDN). We show that the initial feature extraction stage, common to all spatial domain approaches, can be simplified to a Fourier transform with a contourlet-like grouping of coefficients, and saliency detection can be achieved in frequency domain. Specifically, we show that divisive normalization, a model of cortical surround inhibition, can be conducted in frequency domain. Since Fourier coefficients are global in space, we extend to this model by conducting piecewise FDN (PFDN) using overlapping local patches to provide better biological plausibility. Not only do FDN and PFDN outperform current state-of-the-art methods in eye fixation prediction, they are also faster. Speed and simplicity are advantages of our frequency domain approach, and its biological plausibility is the main contribution of our paper.