Outdoor Scene Extrapolation with Hierarchical Generative Cellular Automata
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_ | 1866910483348455424 |
|---|---|
| author | Zhang, Dongsu Williams, Francis Gojcic, Zan Kreis, Karsten Fidler, Sanja Kim, Young Min Kar, Amlan |
| author_facet | Zhang, Dongsu Williams, Francis Gojcic, Zan Kreis, Karsten Fidler, Sanja Kim, Young Min Kar, Amlan |
| contents | We aim to generate fine-grained 3D geometry from large-scale sparse LiDAR scans, abundantly captured by autonomous vehicles (AV). Contrary to prior work on AV scene completion, we aim to extrapolate fine geometry from unlabeled and beyond spatial limits of LiDAR scans, taking a step towards generating realistic, high-resolution simulation-ready 3D street environments. We propose hierarchical Generative Cellular Automata (hGCA), a spatially scalable conditional 3D generative model, which grows geometry recursively with local kernels following, in a coarse-to-fine manner, equipped with a light-weight planner to induce global consistency. Experiments on synthetic scenes show that hGCA generates plausible scene geometry with higher fidelity and completeness compared to state-of-the-art baselines. Our model generalizes strongly from sim-to-real, qualitatively outperforming baselines on the Waymo-open dataset. We also show anecdotal evidence of the ability to create novel objects from real-world geometric cues even when trained on limited synthetic content. More results and details can be found on https://research.nvidia.com/labs/toronto-ai/hGCA/. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_08292 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Outdoor Scene Extrapolation with Hierarchical Generative Cellular Automata Zhang, Dongsu Williams, Francis Gojcic, Zan Kreis, Karsten Fidler, Sanja Kim, Young Min Kar, Amlan Computer Vision and Pattern Recognition We aim to generate fine-grained 3D geometry from large-scale sparse LiDAR scans, abundantly captured by autonomous vehicles (AV). Contrary to prior work on AV scene completion, we aim to extrapolate fine geometry from unlabeled and beyond spatial limits of LiDAR scans, taking a step towards generating realistic, high-resolution simulation-ready 3D street environments. We propose hierarchical Generative Cellular Automata (hGCA), a spatially scalable conditional 3D generative model, which grows geometry recursively with local kernels following, in a coarse-to-fine manner, equipped with a light-weight planner to induce global consistency. Experiments on synthetic scenes show that hGCA generates plausible scene geometry with higher fidelity and completeness compared to state-of-the-art baselines. Our model generalizes strongly from sim-to-real, qualitatively outperforming baselines on the Waymo-open dataset. We also show anecdotal evidence of the ability to create novel objects from real-world geometric cues even when trained on limited synthetic content. More results and details can be found on https://research.nvidia.com/labs/toronto-ai/hGCA/. |
| title | Outdoor Scene Extrapolation with Hierarchical Generative Cellular Automata |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2406.08292 |