DepthGAN: GAN-based depth generation from semantic layouts

Existing GAN-based generative methodsare typically used for semantic image synthesis. Wepose the question of whether GAN-based architecturescan generate plausible depth maps and find thatexisting methods have difficulty in generating depthmaps which reasonably represent 3D scene structuredue to the lack of global geometric correlations.Thus, we propose DepthGAN, a novel method ofgenerating a depth map using a semantic layout asinput to aid construction, and manipulation of wellstructured 3D scene point clouds. Specifically, wefirst build a feature generation model with a cascadeof semantically-aware transformer blocks to obtaindepth features with global structural information.For our semantically aware transformer block, wepropose a mixed attention module and a semanticallyaware layer normalization module to better exploitsemantic consistency for depth features generation.Moreover, we present a novel semantically weighteddepth synthesis module, which generates adaptivedepth intervals for the current scene. We generate thefinal depth map by using a weighted combination ofsemantically aware depth weights for different depthranges. In this manner, we obtain a more accuratedepth map. Extensive experiments on indoor andoutdoor datasets demonstrate that DepthGAN achievessuperior results both quantitatively and visually for thedepth generation task.

Phase-only hologram generation based on integral imaging and its enhancement in depth resolution

We introduce a phase-only hologram generation method based on an integral imaging, and propose an enhancement method in representable depth interval. The computational integral imaging reconstruction method is modified based on optical flow to obtain depth-slice images for the focused objects only. A phaseonly hologram for＇ multiple plane images is generated using the iterative Fresnel transform algorithm. In addition, a division method in hologram plane is proposed for enhancement in the representable minimum depth interval.

Generation of X-ray vortex with ultra-long depth of focus using axial line-focused spiral zone plates

We propose axial line-focused spiral zone plates （ALFSZPs） for generating tightly focused X-ray vortex beams with ultra-long depth of focus （DOF） along the propagation direction. In this typical design, compared with the conventional spiral zone plates （SZPs） under the same numerical aperture （NA）, the DOF of ALFSZPs has been extended to an ultra-length by optimizing the corresponding parameters. Besides, it also exhibits lower side lobes and smaller dark cores in the whole focus volume. The diameters of dark cores increase as the topological charge value increases.