In the past few years, Generative Adversarial Network (GAN) became a prevalent research topic. GAN has ability to generate good quality images that look like natural images from a random vector. In this paper, we follow the basic idea of GAN and propose a novel model for image saliency detection, which is called Supervised Adversarial Networks (SAN). However, different from GAN, the proposed method uses fully supervised learning to learn both G-Network and D-Network by applying class labels of the training set. Moreover, a novel kind of layer call conv-comparison layer is introduced into the D-Network to further improve the saliency performance. Experimental results on Pascal VOC 2012 database show that the SAN model can generate high quality saliency maps for many complicate natural images.
The regular spatial filters comprised of lens and pinhole are essential component in high power laser systems, such as
lasers for inertial confinement fusion, nonlinear optical technology and directed-energy weapon. On the other hand the
pinhole is treated as a bottleneck of high power laser due to harmful plasma created by the focusing beam. In this paper
we present a spatial filter based on angular selectivity of Bragg diffraction grating to avoid the harmful focusing effect in
the traditional pinhole filter. A spatial filter consisted of volume phase gratings in two-pass amplifier cavity were
reported. Two-dimensional filter was proposed by using single Pi-phase-shifted Bragg grating, numerical simulation
results shown that its angular spectrum bandwidth can be less than 160urad. The angular selectivity of photo-thermorefractive
glass and RUGATE film filters, construction stability, thermal stability and the effects of misalignments of
gratings on the diffraction efficiencies under high-pulse-energy laser operating condition are discussed.