Recently, with the advent of deep convolutional neural networks (DCNN), the improvements in visual saliency prediction research are impressive. One possible direction to approach the next improvement is to fully characterize the multiscale saliency-influential factors with a computationally-friendly module in DCNN architectures. In this work, we proposed an end-to-end dilated inception network (DINet) for visual saliency prediction. It captures multi-scale contextual features effectively with very limited extra parameters. Instead of utilizing parallel standard convolutions with different kernel sizes as the existing inception module, our proposed dilated inception module (DIM) uses parallel dilated convolutions with different dilation rates which can significantly reduce the computation load while enriching the diversity of receptive fields in feature maps. Moreover, the performance of our saliency model is further improved by using a set of linear normalization-based probability distribution distance metrics as loss functions. As such, we can formulate saliency prediction as a probability distribution prediction task for global saliency inference instead of a typical pixel-wise regression problem. Experimental results on several challenging saliency benchmark datasets demonstrate that our DINet with proposed loss functions can achieve state-of-the-art performance with shorter inference time.
Well-oriented neodymium doped SnO2 layered nanorod arrays were synthesized by a substrate-free hydrothermal route using sodium stannate and sodium hydroxide at 210 °C. The morphology and phase structure of the Nd-doped SnO2 nanoarrays were investigated by X-ray powder diffraction spectroscopy, scanning electron microscopy, transmission electron microscopy, Raman scattering spectroscopy, X-ray photoelectron spectroscopy and the BET method. The results demonstrated that the Nd-doped SnO2 layered nanorod arrays showed a unique nanostructure combined together with double layered arrays of nanorods with a diameter of 12 nm and a length of several hundred nanometers. The Nd-doped layered SnO2 nanoarrays kept the crystal structure of the bulk SnO2 and possessed more surface defects caused by the Nd ions doped into the SnO2 lattice. The Nd dopant acts as a crystallite growth inhibitor to prevent the growth of SnO2 nanorods. An investigation into the gas-sensing properties indicated that the optimized doping level of 3.0 at% Nd-doped SnO2 layered nanorod arrays exhibited an excellent sensing response toward alcohol at a lower temperature of 260 °C. The enhanced sensor performance was attributed to the higher specific surface area, multi-defect surface structure and the excellent catalytic properties of Nd dopant that is able to increase the amount of active sites on the surface of semiconducting oxides. The Nd-doped SnO2 nanoarray sensors were considered to be a promising candidate for trace alcohol detections in environmental gas monitoring.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.