EarthGen: Generating the World from Top-Down Views
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910599659651072 |
|---|---|
| author | Sharma, Ansh Xiao, Albert Rathi, Praneet Kundu, Rohit Zhai, Albert Shen, Yuan Wang, Shenlong |
| author_facet | Sharma, Ansh Xiao, Albert Rathi, Praneet Kundu, Rohit Zhai, Albert Shen, Yuan Wang, Shenlong |
| contents | In this work, we present a novel method for extensive multi-scale generative terrain modeling. At the core of our model is a cascade of superresolution diffusion models that can be combined to produce consistent images across multiple resolutions. Pairing this concept with a tiled generation method yields a scalable system that can generate thousands of square kilometers of realistic Earth surfaces at high resolution. We evaluate our method on a dataset collected from Bing Maps and show that it outperforms super-resolution baselines on the extreme super-resolution task of 1024x zoom. We also demonstrate its ability to create diverse and coherent scenes via an interactive gigapixel-scale generated map. Finally, we demonstrate how our system can be extended to enable novel content creation applications including controllable world generation and 3D scene generation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2409_01491 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | EarthGen: Generating the World from Top-Down Views Sharma, Ansh Xiao, Albert Rathi, Praneet Kundu, Rohit Zhai, Albert Shen, Yuan Wang, Shenlong Computer Vision and Pattern Recognition Artificial Intelligence J.2; I.4.8 In this work, we present a novel method for extensive multi-scale generative terrain modeling. At the core of our model is a cascade of superresolution diffusion models that can be combined to produce consistent images across multiple resolutions. Pairing this concept with a tiled generation method yields a scalable system that can generate thousands of square kilometers of realistic Earth surfaces at high resolution. We evaluate our method on a dataset collected from Bing Maps and show that it outperforms super-resolution baselines on the extreme super-resolution task of 1024x zoom. We also demonstrate its ability to create diverse and coherent scenes via an interactive gigapixel-scale generated map. Finally, we demonstrate how our system can be extended to enable novel content creation applications including controllable world generation and 3D scene generation. |
| title | EarthGen: Generating the World from Top-Down Views |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence J.2; I.4.8 |
| url | https://arxiv.org/abs/2409.01491 |