 |
ML Newsnew | past | best | rss | submit | about |
| A step towards procedural terrain generation with GANs [pdf] (arxiv.org) | |
| 2 points by blaze33 on Nov 1, 2017 | hide | past | pdf | discuss on HN | |
|
In plain words: Instead of hand-writing rules that turn numbers into hills and valleys, this system learns terrain shapes from NASA satellite pictures and generates new ones. It is an early test showing deep image generators can produce terrain without handcrafted heightmap algorithms. Abstract · A step towards procedural terrain generation with GANsProcedural terrain generation for video games has been traditionally been done with smartly designed but handcrafted algorithms that generate heightmaps. We propose a first step toward the learning and synthesis of these using recent advances in deep generative modelling with openly available satellite imagery from NASA. | |
|
add comment on HN
Also discussed: Nov 2017 (2 points, 0 comments)
|