StyLandGAN: A StyleGAN based Landscape Image Synthesis using Depth-map
This work addresses the need for more expressive input conditions in conditional image synthesis for landscape generation, though it is incremental as it builds on existing StyleGAN frameworks.
The paper tackles the problem of generating landscape images with better control over linear and planar features by using depth maps as input conditions, and demonstrates superior performance in quality, diversity, and depth accuracy compared to modified existing methods.
Despite recent success in conditional image synthesis, prevalent input conditions such as semantics and edges are not clear enough to express `Linear (Ridges)' and `Planar (Scale)' representations. To address this problem, we propose a novel framework StyLandGAN, which synthesizes desired landscape images using a depth map which has higher expressive power. Our StyleLandGAN is extended from the unconditional generation model to accept input conditions. We also propose a '2-phase inference' pipeline which generates diverse depth maps and shifts local parts so that it can easily reflect user's intend. As a comparison, we modified the existing semantic image synthesis models to accept a depth map as well. Experimental results show that our method is superior to existing methods in quality, diversity, and depth-accuracy.