|
|
--- |
|
|
library_name: diffusers |
|
|
pipeline_tag: image-to-image |
|
|
inference: |
|
|
parameters: |
|
|
guidance_scale: 3.5 |
|
|
widget: |
|
|
- src: example_input.jpg |
|
|
text: GenEx Panoramic World Initialization |
|
|
example_title: Panoramic generation from image crop |
|
|
datasets: |
|
|
- TaiMingLu/GenEx-DB-Panorama-World |
|
|
base_model: |
|
|
- black-forest-labs/FLUX.1-Fill-dev |
|
|
license: cc-by-4.0 |
|
|
--- |
|
|
|
|
|
# GenEx-World-Initializer π§π |
|
|
|
|
|
**GenEx World Initializer** is panorama generation pipeline built on top of the [FluxFillPipeline](https://huggingface.co/black-forest-labs/FLUX.1-Fill-dev). |
|
|
|
|
|
It transforms a **single view image** into a **360Β° panoramic image** using vision-conditioned inpainting. |
|
|
|
|
|
- πΌοΈ Input: One image (any size, will be center-cropped to square) |
|
|
- π§ Prompt: Optional text to guide panoramic generation |
|
|
- π― Output: 2048 Γ 1024 equirectangular image |
|
|
- π§© Mask: Uses a fixed panoramic mask |
|
|
|
|
|
|
|
|
|
|
|
 |
|
|
|
|
|
## π¦ Usage |
|
|
|
|
|
```python |
|
|
from diffusers import DiffusionPipeline |
|
|
from PIL import Image |
|
|
import torch |
|
|
|
|
|
pipe = DiffusionPipeline.from_pretrained( |
|
|
"genex-world/World-Initializer-image-to-panorama", |
|
|
custom_pipeline="genex_world_initializer_pipeline", |
|
|
torch_dtype=torch.bfloat16, |
|
|
trust_remote_code=True |
|
|
).to("cuda") |
|
|
|
|
|
# Load your image (any resolution) |
|
|
image = Image.open("example_input.jpg") |
|
|
|
|
|
# Run inference |
|
|
front_view, output = pipe(image=image) |
|
|
output.images[0] |
|
|
``` |
|
|
|
|
|
## π Mask |
|
|
The following mask is used to train the inpainting diffuser and used to inference automatically. |
|
|
 |
|
|
|
|
|
|
|
|
## π§ Requirements |
|
|
|
|
|
```txt |
|
|
diffusers>=0.33.1 |
|
|
transformers |
|
|
numpy |
|
|
pillow |
|
|
sentencepiece |
|
|
``` |
|
|
|
|
|
|
|
|
## β¨ BibTex |
|
|
|
|
|
``` |
|
|
@misc{lu2025genexgeneratingexplorableworld, |
|
|
title={GenEx: Generating an Explorable World}, |
|
|
author={Taiming Lu and Tianmin Shu and Junfei Xiao and Luoxin Ye and Jiahao Wang and Cheng Peng and Chen Wei and Daniel Khashabi and Rama Chellappa and Alan Yuille and Jieneng Chen}, |
|
|
year={2025}, |
|
|
eprint={2412.09624}, |
|
|
archivePrefix={arXiv}, |
|
|
primaryClass={cs.CV}, |
|
|
url={https://arxiv.org/abs/2412.09624}, |
|
|
} |
|
|
``` |