An HVS-Oriented Saliency Map Prediction Model

Visual attention is one of the most significant characteristics for selecting and understanding the outside redundancy world. The nature complex scenes, including larger redundancy and human vision, cannot be processing all information simultaneously because of the information bottleneck. The human visual system mainly focuses on dominant parts of the scenes to reduce the input visual redundancy information. It is commonly known as visual attention prediction or visual saliency map. This paper proposes a new saliency prediction architecture WECSF which inspired by human low-level visual cortex function. The model consists of opponent color channels, wavelet transform, wavelet energy map, and contrast sensitivity function for extracting low-level image features and maximum approximation to the human visual system. The proposed model is evaluated several datasets, including MIT1003, MIT300, TORONTO, SID4VAM and UCF Sports dataset to explain its efficiency. The proposed model results are quantitatively and qualitatively compared to other state-of-the-art salience prediction models. Compared with the baseline model, our model achieved very good performance. Secondly, we also confirmed Fourier/Spectral inspired saliency prediction models has the best prediction scores compared to other start-of-the-art non-neural network and even deep neural network models on the simple image features saliency prediction. Finally, the model also can be applied to spatial-temporal saliency prediction and got better performance.