Human visual attention is subjective and biased according to the personal preference of the viewer, however, current works of saliency detection are general and objective, without counting the factor of the observer. This will make the attention prediction for a particular person not accurate enough. In this work, we propose PANet, a convolutional network that predicts saliency in images with personal preference. The model consists of two streams which share common feature extraction layers, and one stream is responsible for saliency prediction, while the other is adapted from the detection model and used to fit user preference. Experimental results on augmented PASCAL-S and SALICON dataset confirm that PANet can predict saliency areas according to input preference vectors. Comp...[ Read more ]

Human visual attention is subjective and biased according to the personal preference of the viewer, however, current works of saliency detection are general and objective, without counting the factor of the observer. This will make the attention prediction for a particular person not accurate enough. In this work, we propose PANet, a convolutional network that predicts saliency in images with personal preference. The model consists of two streams which share common feature extraction layers, and one stream is responsible for saliency prediction, while the other is adapted from the detection model and used to fit user preference. Experimental results on augmented PASCAL-S and SALICON dataset confirm that PANet can predict saliency areas according to input preference vectors. Compared with other general saliency prediction models, the model with ability of fitting user preference will provide more benefits to either augmented reality (AR) or recommendation applications.