Update model card with metadata and usage instructions
#1
by
nielsr
HF Staff
- opened
README.md
CHANGED
|
@@ -1,16 +1,20 @@
|
|
| 1 |
---
|
| 2 |
-
license: apache-2.0
|
| 3 |
-
language:
|
| 4 |
-
- en
|
| 5 |
base_model:
|
| 6 |
- Wan-AI/Wan2.1-T2V-1.3B
|
| 7 |
- Qwen/Qwen-Image-Edit-2509
|
| 8 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
tags:
|
| 10 |
- video
|
| 11 |
- image
|
| 12 |
- stylization
|
|
|
|
| 13 |
---
|
|
|
|
| 14 |
<div align="center">
|
| 15 |
|
| 16 |
<h1><b>TeleStyle: Content-Preserving Style Transfer in Images and Videos</b></h1>
|
|
@@ -36,19 +40,41 @@ Shiwen Zhang, Xiaoyan Yang, Bojia Zi, Haibin Huang, Chi Zhang, Xuelong Li
|
|
| 36 |
</a>
|
| 37 |
</p>
|
| 38 |
|
|
|
|
| 39 |
|
| 40 |
-
## πNews
|
| 41 |
|
| 42 |
- [2026-01-28]: Released [Code](https://github.com/Tele-AI/TeleStyle), [Model](https://huggingface.co/Tele-AI/TeleStyle), [Paper](http://arxiv.org/abs/2601.20175).
|
| 43 |
|
| 44 |
## How to use
|
| 45 |
|
| 46 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 47 |
|
|
|
|
| 48 |
|
| 49 |
## π Citation
|
| 50 |
|
| 51 |
-
If you find
|
| 52 |
|
| 53 |
```bibtex
|
| 54 |
@article{teleai2026telestyle,
|
|
|
|
| 1 |
---
|
|
|
|
|
|
|
|
|
|
| 2 |
base_model:
|
| 3 |
- Wan-AI/Wan2.1-T2V-1.3B
|
| 4 |
- Qwen/Qwen-Image-Edit-2509
|
| 5 |
+
language:
|
| 6 |
+
- en
|
| 7 |
+
license: apache-2.0
|
| 8 |
+
pipeline_tag: other
|
| 9 |
+
library_name: diffusers
|
| 10 |
+
arxiv: 2601.20175
|
| 11 |
tags:
|
| 12 |
- video
|
| 13 |
- image
|
| 14 |
- stylization
|
| 15 |
+
- style-transfer
|
| 16 |
---
|
| 17 |
+
|
| 18 |
<div align="center">
|
| 19 |
|
| 20 |
<h1><b>TeleStyle: Content-Preserving Style Transfer in Images and Videos</b></h1>
|
|
|
|
| 40 |
</a>
|
| 41 |
</p>
|
| 42 |
|
| 43 |
+
TeleStyle is a lightweight yet effective model for both image and video stylization. Built upon Qwen-Image-Edit, it leverages a Curriculum Continual Learning framework to achieve high-fidelity content preservation and style customization across diverse, in-the-wild style categories.
|
| 44 |
|
| 45 |
+
## π News
|
| 46 |
|
| 47 |
- [2026-01-28]: Released [Code](https://github.com/Tele-AI/TeleStyle), [Model](https://huggingface.co/Tele-AI/TeleStyle), [Paper](http://arxiv.org/abs/2601.20175).
|
| 48 |
|
| 49 |
## How to use
|
| 50 |
|
| 51 |
+
### 1. Installation
|
| 52 |
+
|
| 53 |
+
```bash
|
| 54 |
+
pip install -r requirements.txt
|
| 55 |
+
```
|
| 56 |
+
|
| 57 |
+
### 2. Inference
|
| 58 |
+
|
| 59 |
+
We provide inference scripts for running TeleStyle on demo inputs for each task:
|
| 60 |
+
|
| 61 |
+
#### Image Stylization
|
| 62 |
+
To generate a stylized image using a reference style image and a content image:
|
| 63 |
+
```bash
|
| 64 |
+
python telestyleimage_inference.py --image_path assets/example/0.png --style_path videos/1.png --output_path results/image.png
|
| 65 |
+
```
|
| 66 |
+
|
| 67 |
+
#### Video Stylization
|
| 68 |
+
To generate a stylized video using a stylized first frame and a content video:
|
| 69 |
+
```bash
|
| 70 |
+
python telestylevideo_inference.py --video_path assets/example/1.mp4 --style_path assets/example/1-0.png --output_path results/video.mp4
|
| 71 |
+
```
|
| 72 |
|
| 73 |
+
For more details, please refer to the [**π GitHub**](https://github.com/Tele-AI/TeleStyle) repository.
|
| 74 |
|
| 75 |
## π Citation
|
| 76 |
|
| 77 |
+
If you find TeleStyle useful for your research and applications, please cite using this BibTeX:
|
| 78 |
|
| 79 |
```bibtex
|
| 80 |
@article{teleai2026telestyle,
|