@endraven4
Profilo
Registrato: 1 anno, 9 mesi fa
ICCV 2022 Open Access Repository Zekun Hao, Arun Mallya, Serge Belongie Ming-Yu Liu, Zekun Hao; Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) 2021 (pp. 14072-14082 Abstract We present GANcraft, a neural unsupervised rendering framework for generating photos of realistic images of huge 3D block worlds like the ones created in Minecraft. Our approach uses an underlying semantic block world as an input, where each block is given a semantic name such as dirt, grass, or water. We model the world as a continuous volumetric function and train our model to render view-consistent , photorealistic images for a user-controlled camera. We designed a training technique that is based on adversarial and pseudo-ground truth training in the absence images from the block world. This stands in contrast to prior research on neural rendering to view synthesis, which relies on ground truth images to calculate scene geometry and view-dependent appearance. GANcraft lets users control the semantics of the scene as well as output style. lalalalal Results from experiments compared to strong baselines show the effectiveness of GANcraft in this new task of photorealistic 3D block synthesizing.
Sito web: http://cardraven1.jigsy.com/entries/general/How-To-Download-Minecraft-Minecraft
Forum
Topic aperti: 0
Risposte create: 0
Ruolo forum: Partecipante