|
--- |
|
tags: |
|
- image-to-image |
|
- lora |
|
- diffusers |
|
- template:diffusion-lora |
|
widget: |
|
- output: |
|
url: images/_ki-bY_sewZCIApJgDYQv.png |
|
text: Replace the face with Nicolas Cage's face |
|
- output: |
|
url: images/8VjJQRIuk6hOKdVSRQHaN.png |
|
text: Replace the face with Nicolas Cage's face |
|
- output: |
|
url: images/cO-qVnU-o79Zo94LMyF2w.png |
|
text: Replace the face with Nicolas Cage's face |
|
- output: |
|
url: images/HBefBm8iIFcH--IMcipi0.png |
|
text: Replace the face with Nicolas Cage's face |
|
- output: |
|
url: images/leLq1_DkeopjjKybaqKY2.png |
|
text: Replace the face with Nicolas Cage's face |
|
base_model: black-forest-labs/FLUX.1-Kontext-dev |
|
instance_prompt: null |
|
license: mit |
|
--- |
|
# FLUX.1-Kontext-dev-cage-ify |
|
|
|
<Gallery /> |
|
|
|
## Model description |
|
|
|
# ๐ญ Cage-ify |
|
|
|
[](https://fal.ai/models/fal-ai/image-editing/cage-ify) |
|
[](https://opensource.org/licenses/MIT) |
|
[](https://twitter.com/ThytuVDM) |
|
|
|
> Transform any face into Nicolas Cage's face while maintaining the original expression, lighting, and pose. |
|
|
|
## ๐ฏ What is Cage-ify? |
|
|
|
Cage-ify is a custom LoRA fine-tune of FLUX.1 Kontext that can replace any human face with Nicolas Cage's face while preserving the original facial expression, lighting conditions, and head positioning. |
|
|
|
### โจ Key Features |
|
|
|
- **๐ญ Perfect Face Replacement**: Seamlessly swaps any face with Nicolas Cage |
|
- **๐จ Expression Matching**: Maintains original facial expressions |
|
- **๐ก Lighting Preservation**: Keeps original lighting and shadows |
|
- **๐ Pose Retention**: Preserves head position and angle |
|
- **๐ Universal Compatibility**: Works on any photo with a visible face |
|
|
|
## ๐ Quick Start |
|
|
|
### Try Online (Recommended) |
|
[**๐ฎ Try Cage-ify Live โ**](https://fal.ai/models/fal-ai/image-editing/cage-ify) |
|
|
|
### API Usage |
|
```python |
|
import fal_client |
|
|
|
def cage_ify_image(image_url): |
|
result = fal_client.submit( |
|
"fal-ai/flux-kontext-lora", |
|
arguments={ |
|
"image_url": image_url, |
|
"prompt": "Transform the face into Nicolas Cage's face", |
|
"lora_path": "your-lora-path-here" |
|
} |
|
) |
|
return result |
|
``` |
|
|
|
## ๐ ๏ธ How It Works |
|
|
|
### Training Process |
|
1. **Data Collection**: Gathered 100+ high-quality Nicolas Cage images |
|
2. **Reverse Engineering**: Used FLUX.1 Kontext to transform Cage faces into diverse other faces |
|
3. **Dataset Creation**: Flipped the pairs to create "other face โ Cage" training data |
|
4. **LoRA Training**: Fine-tuned FLUX.1 Kontext [dev] using the dataset |
|
|
|
## ๐จ Use Cases |
|
|
|
If you really need a reason to why we should Cage-ify everything, here are some: |
|
- **๐ฌ Movie Poster Parodies**: Cage-ify your favorite films |
|
- **๐จโ๐ฉโ๐งโ๐ฆ Family Photos**: Make everyone in your family look like Nicolas Cage |
|
- **๐จ Art Projects**: Creative digital art and collages |
|
- **๐ Our True Leader**: Spray the words |
|
|
|
## ๐ Acknowledgments |
|
|
|
- **Black Forest Labs** for creating FLUX.1 Kontext |
|
- **fal.ai** for providing training and inference infrastructure |
|
- **Nicolas Cage** for being the One True God of faces |
|
- **The Internet** for existing |
|
|
|
## ๐ Contact |
|
|
|
- **Twitter**: [@ThytuVDM](https://twitter.com/ThytuVDM) |
|
- **Issues**: [GitHub Issues](https://github.com/Thytu/cage-ify/issues) |
|
|
|
--- |
|
|
|
<div align="center"> |
|
|
|
**โญ Star this repo if Cage-ify brought joy to your life! โญ** |
|
|
|
*"I'm gonna steal the Declaration of Independence... and every face on the internet."* - Nicolas Cage (probably) |
|
|
|
</div> |
|
|
|
|
|
|
|
## Download model |
|
|
|
|
|
[Download](/Thytu/FLUX.1-Kontext-dev-cage-ify/tree/main) them in the Files & versions tab. |
|
|