Incorporating visual field characteristics into a saliency map

Abstract

Characteristics of the human visual field are well known to be different in central (fovea) and peripheral areas. Existing computational models of visual saliency, however, do not take into account this biological evidence. The existing models compute visual saliency uniformly over the retina and, thus, have difficulty in accurately predicting the next gaze (fixation) point. This paper proposes to incorporate human visual field characteristics into visual saliency, and presents a computational model for producing such a saliency map. Our model integrates image features obtained by bottom-up computation in such a way that weights for the integration depend on the distance from the current gaze point where the weights are optimally learned using actual saccade data. The experimental results using a large number of fixation/saccade data with wide viewing angles demonstrate the advantage of our saliency map, showing that it can accurately predict the point where one looks next.

Psychophysical evaluation of individual low-level feature influences on visual attention

“…The temporal evolution of fixation and saccade behavior reveal distinct patterns of eye movements upon viewing time , confirming the evidence that visual attention is an active process and its modeling involving temporality requires further investigation. Scanpath prediction could allow the reproduction of the aforementioned effects, regarding in that aspect both bottom-up and top-down processing of visual features that distinctively guide visual attention [Boccignone and Ferraro, 2004] [Kubota et al, 2012] [Chang et al, 2014] [LeMeur and Liu, 2015] [Aboudib et al, 2015] [Adeli et al, 2016] [Wang et al, 2016] [Wloka et al, 2017] [White et al, 2017. In that aspect, as saliency decreases over time, saliency evaluation measures should be done in that line.…”

Computational Models of Human Visual Attention and Their Implementations: A Survey

“…Recently the models have been used to boost some computer vision and pattern recognition techniques such as object detection [113], [124], [125], object recognition [126]- [131], action recognition [132], [133], segmentation [37], [114], [115], [134], [135] and background subtraction [136]. Besides, specific applications include video summarization [137] and compression [138], scene understanding [139]- [141], computer-human interaction [98], [142]- [147], robotics [132], [148]- [150], and driver assistance [151], [152]. The potential of the visual attention models that are capable of extracting important regions promises their contributions to many other domains.…”

Influence of stimulus and viewing task types on a learning-based visual saliency model

“…Using the dataset described above, visual saliency models are learned according to [Kubota et al 2012]. Unlike other learningbased models, their model takes into account non-uniformity of sensitivity within a field of view by using different weights depend- ing on the distance from a current fixation position.…”

“…Zhao et al [2011] proposed a model trained with features used in Itti et al's model [1998] by the least squares method. Kubota et al [2012] improved the learning-based saliency model by incorporating characteristics of the human visual field.…”