Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,91 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: apache-2.0
|
| 3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
---
|
| 4 |
+
|
| 5 |
+
<p align="center">
|
| 6 |
+
<h2 align="center">SceneDesigner: Controllable Multi-Object Image Generation with 9-DoF Pose Manipulation</h2>
|
| 7 |
+
<p align="center">
|
| 8 |
+
<a href="https://github.com/qqzy/"><strong>Zhenyuan Qin<sup>*</sup></strong></a>
|
| 9 |
+
·
|
| 10 |
+
<a href="https://github.com/xinchengshuai/"><strong>Xincheng Shuai<sup>*</sup></strong></a>
|
| 11 |
+
·
|
| 12 |
+
<a href="https://henghuiding.com/"><strong>Henghui Ding </strong><sup>†</sup></a>
|
| 13 |
+
</p>
|
| 14 |
+
|
| 15 |
+
<p align="center">
|
| 16 |
+
Fudan University
|
| 17 |
+
</p>
|
| 18 |
+
<p align="center">
|
| 19 |
+
<a href="https://arxiv.org/abs/2511.16666"><img src="https://img.shields.io/static/v1?label=Paper&message=2511.16666&color=red&logo=arxiv"></a>
|
| 20 |
+
<a href="https://henghuiding.com/SceneDesigner/"><img src="https://img.shields.io/static/v1?label=Project%20Page&message=Github&color=blue&logo=github-pages"></a>
|
| 21 |
+
<a href="https://huggingface.co/datasets/FudanCVL/ObjectPose9D"><img src="https://img.shields.io/badge/🤗_HuggingFace-Dataset-ffbd45.svg" alt="HuggingFace">
|
| 22 |
+
</p>
|
| 23 |
+
|
| 24 |
+
|
| 25 |
+
|
| 26 |
+
|
| 27 |
+
|
| 28 |
+
<!-- ## 🎯 Introduction -->
|
| 29 |
+
|
| 30 |
+
|
| 31 |
+

|
| 32 |
+
|
| 33 |
+
## ⚙️ Quick Start
|
| 34 |
+
|
| 35 |
+
### 1. Installation
|
| 36 |
+
|
| 37 |
+
1. Install Python environment (recommended to use uv)
|
| 38 |
+
```bash
|
| 39 |
+
uv sync
|
| 40 |
+
```
|
| 41 |
+
Or alternatively:
|
| 42 |
+
```bash
|
| 43 |
+
pip install -r requirements.txt
|
| 44 |
+
```
|
| 45 |
+
|
| 46 |
+
2. Install Blender environment
|
| 47 |
+
```bash
|
| 48 |
+
cd render
|
| 49 |
+
python install.py
|
| 50 |
+
```
|
| 51 |
+
|
| 52 |
+
If the automatic installation script fails, you can install manually:
|
| 53 |
+
* First download [Blender](https://download.blender.org/release/Blender4.2/) and extract it to the `./render` directory
|
| 54 |
+
* Then locate the Blender Python path and install the Python dependencies for Blender, for example:
|
| 55 |
+
```bash
|
| 56 |
+
cd render
|
| 57 |
+
blender-4.2.8-linux-x64/4.2/python/bin/python3.11 -m pip install -r blender_requirements.txt
|
| 58 |
+
```
|
| 59 |
+
|
| 60 |
+
### 2. Download Checkpoints
|
| 61 |
+
|
| 62 |
+
1. Download the [SceneDesigner](https://huggingface.co/FudanCVL/SceneDesigner) weights to the `checkpoints` directory
|
| 63 |
+
2. Download the [Stable Diffusion 3.5](https://huggingface.co/stabilityai/stable-diffusion-3.5-medium) base model weights to the `checkpoints` directory
|
| 64 |
+
|
| 65 |
+
### 3. Run Demo
|
| 66 |
+
|
| 67 |
+
Launch the Gradio app:
|
| 68 |
+
```bash
|
| 69 |
+
python app.py \
|
| 70 |
+
--blender_path render/blender/blender \
|
| 71 |
+
--device cuda:0 \
|
| 72 |
+
--port 7861
|
| 73 |
+
```
|
| 74 |
+
|
| 75 |
+
- Adjust the 9D pose of the cube in the **Cube Controls** panel
|
| 76 |
+
- Enter text prompts in the **Generation Config** panel and click the **Generate Images** button to create images
|
| 77 |
+
|
| 78 |
+
|
| 79 |
+
|
| 80 |
+
## ✒️ Citation
|
| 81 |
+
|
| 82 |
+
If you find our work useful for your research and applications, please kindly cite using this BibTeX:
|
| 83 |
+
|
| 84 |
+
```latex
|
| 85 |
+
@inproceedings{SceneDesigner,
|
| 86 |
+
title={SceneDesigner: Controllable Multi-Object Image Generation with 9-DoF Pose Manipulation},
|
| 87 |
+
author={Qin, Zhenyuan and Shuai, Xincheng and Ding, Henghui},
|
| 88 |
+
booktitle={NeurIPS},
|
| 89 |
+
year={2025}
|
| 90 |
+
}
|
| 91 |
+
```
|