Draw to Photo | Draw2Photo

#flux#text-to-image#lora#diffusers#fal
Base Model

black-forest-labs/FLUX.1-Kontext-dev

Instance Prompt

make it real

Pipeline Tag

image-to-image

Trigger Words

You should use make it real to trigger the image generation.

Examples

Input Sketch
Output Photo
Input 1
Output 1
Input Sketch
Output Photo
Input 2
Output 2
Input Sketch
Output Photo
Input 3
Output 3
Input Sketch
Output Photo
Input 4
Output 4
Input Sketch
Output Photo
Input 5
Output 5
Input Sketch
Output Photo
Input 6
Output 6
Input Sketch
Output Photo
Input 7
Output 7
Input Sketch
Output Photo
Input 8
Output 8
Input Sketch
Output Photo
Input 9
Output 9

How to use

  1. Prepare your inputs: You need a drawing and a face photo. Put the face photo on the right top of the sketch.
  2. Add trigger word: Include make it real in your prompt.
  3. Tune Settings: Adjust LoRA weight (recommended 1.0).

Try it Online

We have built a demo space to try it easily.

Visit Demo Space

Try it Locally

git clone https://huggingface.co/spaces/ovi054/Draw2Photo cd Draw2Photo pip install -r requirements.txt python app.py

Download Model

Weights for this model are available in Safetensors format.

Download Files & Versions

Training

Training was done using fal.ai/models/fal-ai/flux-kontext-trainer.

FAQ

What is Draw to Photo?

It is an AI demo tool that transforms your simple sketches into realistic portraits by combining drawn poses with uploaded face photos.

How to get best results?

Place face photo at top-right of sketch. Ensure face is clear. Use trigger word "make it real". LoRA weight 1.0 recommended.

Are my photos saved?

Depends on the hosting service. Local runs save locally. If on Hugging Face Space, check their privacy policy.

Can I use colored drawings?

The model is trained mainly on simple sketches. Colored drawings might yield unpredictable results; simple line art is recommended.