Center bias outperforms image salience but not semantics in accounting for attention during scene viewing
- PDF / 1,447,394 Bytes
- 10 Pages / 595.224 x 790.955 pts Page_size
- 59 Downloads / 185 Views
Center bias outperforms image salience but not semantics in accounting for attention during scene viewing Taylor R. Hayes1 · John M. Henderson1,2 © The Psychonomic Society, Inc. 2019
Abstract How do we determine where to focus our attention in real-world scenes? Image saliency theory proposes that our attention is ‘pulled’ to scene regions that differ in low-level image features. However, models that formalize image saliency theory often contain significant scene-independent spatial biases. In the present studies, three different viewing tasks were used to evaluate whether image saliency models account for variance in scene fixation density based primarily on scene-dependent, low-level feature contrast, or on their scene-independent spatial biases. For comparison, fixation density was also compared to semantic feature maps (Meaning Maps; Henderson & Hayes, Nature Human Behaviour, 1, 743–747, 2017) that were generated using human ratings of isolated scene patches. The squared correlations (R 2 ) between scene fixation density and each image saliency model’s center bias, each full image saliency model, and meaning maps were computed. The results showed that in tasks that produced observer center bias, the image saliency models on average explained 23% less variance in scene fixation density than their center biases alone. In comparison, meaning maps explained on average 10% more variance than center bias alone. We conclude that image saliency theory generalizes poorly to real-world scenes. Keywords Scene perception · Center bias · Saliency · Semantics · Meaning map Real-world visual scenes are too complex to be taken in all at once (Tsotsos, 1991; Henderson, 2003). To cope with this complexity, our visual system uses a divide-and-conquer strategy by shifting our attention to different smaller subregions of the scene over time (Findlay & Gilchrist, 2003; Henderson & Hollingworth, 1999; Hayhoe & Ballard, 2005). This solution leads to a fundamental question in cognitive science: How do we determine where to focus our attention in complex, real-world scenes? One of the most influential answers to this question has been visual salience. Image salience theory proposes that our attention is ‘pulled’ to visually salient locations that differ from their surrounding regions in semantically uninterpreted image features like color, orientation, and luminance (Itti & Koch, 2001). For example, a search array that contains a single red line among an array of green lines stands out and draws our attention (Treisman & Gelade 1980; Taylor R. Hayes
[email protected] 1
Center for Mind and Brain, University of California, Davis, CA, USA
2
Department of Psychology, University of California, Davis, CA, USA
Wolfe, Cave, & Franzel, 1989; Wolfe 1994). The idea of visual salience has been incorporated into many influential theories of attention (Wolfe & Horowitz, 2017; Itti & Koch, 2001; Wolfe et al., 1989; Treisman & Gelade, 1980) and formalized in various computational image saliency models (Itti, Koch, & Niebur, 1998; Harel,
Data Loading...
