Commit
•
d69bd7b
1
Parent(s):
2bd4d95
Update README.md
Browse files
README.md
CHANGED
|
@@ -28,6 +28,14 @@ Objaverse-XL is over an order of magnitude larger and much more diverse!
|
|
| 28 |
|
| 29 |
<img src="https://github.com/allenai/objaverse-rendering/assets/28768645/43833dd3-ec97-4a3d-8782-00a6aea584b4">
|
| 30 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 31 |
## Citation
|
| 32 |
|
| 33 |
To cite Objaverse-XL, please cite our [📝 arXiv](https://arxiv.org/abs/2307.05663) paper with the following BibTeX entry:
|
|
|
|
| 28 |
|
| 29 |
<img src="https://github.com/allenai/objaverse-rendering/assets/28768645/43833dd3-ec97-4a3d-8782-00a6aea584b4">
|
| 30 |
|
| 31 |
+
## Unlocking Generalization
|
| 32 |
+
|
| 33 |
+
Compared to the original Zero123 model, Zero123-XL improves remarkably in 0-shot generalization abilities, even being able to perform novel view synthesis on sketches, cartoons, and people!
|
| 34 |
+
|
| 35 |
+
A ton more examples in the [📝 paper](https://arxiv.org/abs/2307.05663) :)
|
| 36 |
+
|
| 37 |
+
<img src="https://github.com/allenai/objaverse-rendering/assets/28768645/8470e4df-e39d-444b-9871-58fbee4b87fd">
|
| 38 |
+
|
| 39 |
## Citation
|
| 40 |
|
| 41 |
To cite Objaverse-XL, please cite our [📝 arXiv](https://arxiv.org/abs/2307.05663) paper with the following BibTeX entry:
|