To compute a probabilistic map of an image to predict the actual human eye gaze patterns
To generate masks that matches the annotated silhouettes of salient objects
Images from the validation set of PASCAL VOC 2010
Each image is manually segmented for salient object annotation
Eye tracking from multiple subjects during a two-second free viewing
Salient objects selected by multiple subjects using the segmentations
The definition of Salient Objects is an intrinsic property of the image, which can be reliably perceived among human subjects.
Unlike fixation datasets, the most widely used salient object segmentation dataset is heavily biased.
There exists a strong correlation between fixations and salient objects, which can be used to improve salient object segmentation.
We propose a salient object segmentation model by combining object proposal and fixation prediction